Volume II/4. The Student Loans Scheme in Thailand: A Review and Recommendations for Efficient and Equitable Functioning of the Scheme

Transcription

1 Volume II/4 The Student Loans Scheme in Thailand: A Review and Recommendations for Efficient and Equitable Functioning of the Scheme Prepared for UNESCO-Bangkok as part of the Asian Development Bank Social Sector Program Loan in the framework of the Education Management and Finance Study Project TA2996-THA Written by Adrian Ziderman Bar Ilan University, Israel July, 1999

2 Acknowledgements This report was written by Dr. Adrian Ziderman. He headed a team of national consultants from the UNESCO Principal Regional Office for Asia and the Pacific (PROAP), which carried out the student loans scheme review between July 1998 and March 1999 in the framework of the ADB-funded Education Management and Financing Study (T.A THA). The review benefited from substantive and administrative support provided by the Office of the National Education Commission (ONEC). The following individuals were involved in this study: Consultant Team Dr. Adrian Ziderman, UNESCO international consultant Professor of Economics and of Education, Bar-Ilan University, Ramat Gan, Israel Editor, International Journal of Manpower Dr. Nonglak Wiratchai, UNESCO national consultant Assistant Professor and Head of the Department of Educational Research Chulalongkorn University, Bangkok Dr. Supang Chantavanich, UNESCO national consultant Associate Professor and Director of the Asian Research Center for Migration, at the Institute for Asian Studies, Chulalongkorn University, Bangkok Dr. Kitti Limskul, UNESCO national consultant Director of Economic Modeling and Forecasting Program Chulalongkorn University, Bangkok ONEC Officials and Researchers Ms Walairat Asaves, Education Officer, ONEC Ms Taweeporn Boonanich, Data Processing Officer, ONEC Ms Rungkarn Suksamai, Education Officer, ONEC Ms Achariyah Kohtbantau, Graduate Student, University of Oregon, USA Mr. Nonthipoth, Graduate Student, Thammasat University, Bangkok ii

3 Table of Contents A. Executive Summary Introduction International Experience Major Findings of the Review Coverage Targeting Finance Towards Reform of the Student Loans Scheme Organizational Reform: Improving Targeting and Horizontal Equity Finance Reform Improving Performance: Administration and Process B. The Student Loan Scheme in Thailand: Present Performance and Future Reform Objectives and context of the review Terms of Reference The work program Plan of the review International experience with student loans The main features of the Student Loans Scheme Objectives and expectations Loans Scheme: central administrative bodies and responsibilities National Loans Scheme budget Loans budget allocation Loan distribution Concluding remarks SLS Breakdown and coverage by type of institution Number of borrowers and average loan size Coverage Review of Loans Scheme Administration Evaluation: process versus outcomes Methodology Administration Problems: Input and Process Targeting and horizontal equity Objectives The Reach of the Loans Scheme Demand for Student Loans Horizontal equity Case Study: MOE Provincial Loans Budget Allocation iii

4 7. Financial analysis Alternative objectives and policies The individual loan account Loan recovery with default and administration costs (the Subsidy-Loan Ratio) Loans Scheme Viability: Revolving Fund or Open-ended Commitment Towards the future: an agenda for reform Organizational reform: improving targeting and horizontal equity Improving financial performance Improving performance: administration and process Reform of the loans scheme and ongoing plans for educational reform Bibliography Appendix iv

5 Table of Tables Table 1 Recovery from student loans schemes, selected countries Table 2 Student Loan Scheme: planned and actual budget, number of loan recipients, Table 3 Number of borrowers and total loans expenditure: planned and actual Table 4 Student loans fund allocation for new recipients, Table 5 Number of borrowers and total loans distribution, Table 6 Loans budget allocation, MOE Table 7 Loans maxima, Table 8 Student Loans Scheme: Number of new and old recipients Table 9 Student Loans Scheme: Average loan size (Baht) Table 10 Student loans scheme coverage Table 11 Matrix of loan administration problems classified by input and process Table 12 Attendance at Meetings of Student Loans Committees, Table 13 Reach of the Loans Scheme: Upper Secondary Level of Schooling Table 14 Number of Poor Youth Aged in Poor Households, by School Attendance, Table 15 Annual Household Income of Loan Recipients in Selected Provinces and Institutions (per cent) Table 16 Possible Consequences of Lack of Loan for Recipients Table17 Average Loan Size, Recipients in MUA Demonstration Schools, Table 18 Loan Distribution, by Parental Income, Khon Kaen University and North-Eastern College (Khon Kaen), Table 19 Maximum loan ceilings, RIT (Khon Kaen), Table 20 Number of loan recipients at Thammasat University, by gender and family income (1997) Table 21 Correlation matrix of five target measures Table 22 Bivariate correlations between loan allocation and target measures Table 23 Hidden grants and repayment ratio on selected student loans programs: alternative discount rates (per cent) Table 24 Loan repayment as a percentage of annual income (in terms of present values at a 6% discount rate) Table 25 Effect on default of the repayment rates Table 26 Net recovery ratio on loans (with 5% default and alternative administration costs) Table 27 Repayment ratio: The effect of alternative changes on loan repayment conditions, undergraduate programs (per cent) v

6 Table28 Recovery ratios on loans: the effect of changes in repayment conditions at alternative levels of administration costs undergraduate programs (per cent) Table 29 Present values of repayment burden (6% discount rate): loan repayment as a percentage of annual income, allocation repayment condition (per cent) vi

7 Table of Figures Figure 1 Student Loans Scheme Coverage, Figure 2 Student Loan Fund Allocation... 5 Figure 3 Reach of the Scheme... 7 Figure 4 Alternative Scenarios for Loans Allocation Reform Figure 5 Average Loan Size (new borrowers) Figure 6 Student loan budget allocation Figure 7 Student loans scheme coverage Figure 8 Reach of the Scheme Figure 9 Target borrowing populations, schools, Figure 10 Percentage of borrowers by income group, Figure 11 Loan ceilings, RIT Khon Kaen, Figure 12 Scatter diagram of average loan per student (LS) and poor students ages as a percentage of population (POORST) ( ) Figure 13 Alternative scenarios for loans allocation reform vii

8 Appendix Tables Table 1 Student loans: total loan allocation, number of recipients and average loan size, alternative education programs Table 2 Student loans scheme: average loan size, old and new recipients, Table 3 Number of loan recipients, total loans and average size of loan, by level and year of schooling, Table 4 Institutions visited Table 5 Number and percentage of loan recipients by income group and institution, selected universities, Table 6 List of provinces in each of the four allocation segments (POORST)/target groups (LS) in Figure Table 7 List of the provinces in each of four allocation (POORST)/target groups (RS) Table 8 Annual loan repayment burden, upper secondary courses, Table 9 by gender (per cent) Annual loan repayment burden, undergraduate courses, by gender (per cent) Table 10 Annual loan repayment burden, upper secondary school plus undergraduate courses, by gender (per cent) Table 11 Early repayment to the Student Loan Scheme (April 1997-September 1998) Table 12 Administration Costs Table 13 Net recovery ratios on loans: the effect of changes in repayment conditions at alternative levels of administration costs, undergraduate programs (per cent) Table 14 Annual loan repayment burden, undergraduate courses, by gender: alternative repayment conditions (per cent) Table 15 Present values of repayment burden (6% discount rate): loan repayment as a percentage of annual income, allocation repayment conditions (per cent) viii

9 List of Abbreviations ADB GPPPC KTB LEAs MOE MOF MUA NEC NESDB ONEC RIT SES SLB SLS SLSC Asian Development Bank Gross Provincial Product per capita Krung Thai Bank Local Education Authorities Ministry of Education Ministry of Finance Ministry of University Affairs National Education Commission National Economic and Social Development Board Office of the National Education Commission Rajamamgala Institute of Technology Social and Economic Survey (of the NESDB) Student Loans Bureau Student Loans Scheme Student Loans Scheme Committee ix

10 A. Executive Summary 1 1. Introduction a. The decision to establish a student loans scheme (SLS) in Thailand was made by the Cabinet on March ; the scheme came into effect in the academic year 1996.The scheme operated under this Cabinet resolution until the promulgation of the Student Loans Scheme Act in March Now only in its third academic year of operation, the accomplishments of the SLS are clearly considerable. From a general perspective, the scheme has been a success. The SLS currently dispenses a total loans budget of over 20 billion Baht, providing loans to some 650,000 students at the secondary and tertiary levels of education. A comprehensive administrative structure has been put in place to allocate the loans budgets within the Ministries of Education and University Affairs and the Krung Thai Bank has established a national financial network to service the banking aspects of the scheme. b. At a more fundamental level, however, the general efficacy of the scheme remains rather problematic, particularly in relation to the targeting of the poor and to its financial sustainability. These problems seem to stem from the somewhat hasty manner in which the scheme was set up, albeit with the laudable aim of providing assistance to poor students with a minimum of delay. The scheme closely mirrored the proposals outlined in a background document prepared for the cabinet by the Ministry of Finance (March 1995); this paper was lacking in two main respects. First, its proposals for the administrative structure for the scheme paid no attention to the need to establish new organizational structures directly aimed at achieving the over riding objective of well tuned loans targeting. Rather, it saw the issue of loans distribution in more narrowly traditional terms, as one relating to top-down budget allocation, which was not suitable for the social objectives underlying the scheme. Secondly, the financial analysis and forecasts provided in the paper were weak, resulting in an unduly optimistic prognosis of the expected impact of the project and its future budgetary burden. c. The central weaknesses of the scheme arise from poor design, in turn the result of inadequate planning and hasty implementation. Yet the new Student Loans Scheme Act closely echoes practice under the cabinet resolution; little advantage has been taken of the accumulated experience of two years of operation to both identify existing shortcomings of the scheme and to fashion the required changes in the new Act. The Act has been on the statute book for over a year; yet the SLS continues to operate without the direction of a fulltime manager nor the necessary inputs of senior technical and professional staff. In practice, this has led to weak overall direction and control of the scheme, with minimal forward planning, monitoring or evaluation. Thus the SLS has been largely reactive, dealing with issues as they arise on an ad hoc basis rather than adopting a desirable, proactive stance and displaying strong direction from the center. 1 A note on terminology: we use the term loans allocation to refer to the institutional division of loans budgets at various levels in the organization hierarchy; loans distribution refers to the division of a given loans budget amongst loans applicants. 1

11 2. International Experience a. The Thai student loans scheme may be viewed against the background of the accumulated experience from student loans that are in place in over 50 countries throughout the world. In the region, loans schemes have operated in a number of countries, including China, Hong Kong, Indonesia, Japan, Korea, Malaysia, the Philippines, Singapore and Australia. The economic logic underlying student loans schemes is unassailable. Since additional educational qualifications lead to enhanced labor market earnings, loans enable students to delay payments for education (whether for education-related costs or living expenses) until they are earning the enhanced earnings that the additional education has made possible. b. However, the central objective of loans scheme varies from country to country. The strongest case for the introduction of a loans scheme can be made where its purpose is to facilitate greater cost recovery and lower public expenditure by enabling educational institutions to raise tuition fees. Other schemes are aimed at easing the financial burdens of students (including the non-poor) during their study years. In contrast, the Thai scheme is aimed essentially at increasing the educational participation of disadvantaged students. c. Whatever the underlying purpose (and many schemes combine more than one objective), few student loans schemes have proved to be financially viable, particularly in developing countries, in the sense that loan recovery in practice is typically between per cent and very seldom reaches 70 per cent. This has been the case particularly where a combination of highly subsidized interest charges on loans, default on loan repayments and high administration costs have resulted in low average recovery ratios (repayments in relation to loan size). Low loan recovery ratios may be justified where the aims of the scheme are dominantly social, aimed at increasing the educational access of the poor. However, in these cases, as with the Thai loans scheme, the question is whether the level of subsidy (the inverse of loan recovery) is justified in terms of the social benefits received. This, in turn, will depend on how well the subsidized loans are targeted to reach the disadvantaged, especially at the lower tier. d. In virtually every case, student loans schemes are confined to tertiary education. International experience shows that the accepted mechanism for aiding poor youngsters at secondary school is through the provision of scholarships (money grants) directly targeted to poor students and their families. In the much quoted loans scheme in Sweden, loans are provided for secondary school students; however, even in that case, loans constitute only a small proportion of the total aid package, the bulk of which consists of grants. The provision of loans for youngsters at high school in the Thai case departs markedly from accepted international practice. 2

12 3. Major Findings of the Review A broad program of background research was carried out, in order both to identify any major shortcomings of the present scheme and to underpin recommendations for reform. Indeed, in thinking towards reform of the current student loans scheme, our findings relating to three major issues were seen to be of central importance. The first concerns the unusually broad and growing coverage of the scheme, with secondary school youngsters constituting the vast majority of borrowers; this is coupled with falling loan budgets, in real terms. Secondly, there is a consensus that the central guiding principle underlying the SLS is to satisfy the twin objectives of reaching needy students (targeting) and providing equality of treatment to borrowers in the distribution of individual loans (horizontal equity); how well are these objectives secured, in practice? In particular, is the organizational structure of the SLS and its administrative procedures conducive to targeting and equity? Third, the financial implications of the loans scheme and its sustainability over the longer term need to be examined carefully; it appears that no financial appraisal of the scheme, neither at the stage of planning the SLS nor subsequently, has been undertaken until now. These three issues are central to any consideration of the efficacy of present SLS arrangements and to the fashioning of directions for reform. 3.1 Coverage a. Virtually all student loans schemes in other countries are confined to tertiary level education. The SLS is unusual in including also borrowers at the secondary school level; this group now constitutes a sizeable majority of borrowers, accounts for over 60 per cent of the total loans budget, and imposes a considerable administrative burden on the education system. b. Planning documents envisaged a build-up of loan recipients in the early years, reaching a steady state of 300,000 borrowers in the fourth year of the scheme. These early plans have since been overtaken by events. The 300,000 borrowers target was exceeded in the second year of the scheme; currently the target number of borrowers is outstripped by a factor of 2.3. The scheme was planned originally to cover some 9 per cent of enrollments at each education level; the coverage of the scheme (total number of current loan recipients divided by the total student enrollment) is far in excess of this. By 1998, overall coverage exceeded 20% of all enrollments, nearly 25% for secondary schooling and 15% at the tertiary level (Figure 1) c. Coverage at MOE institutions is high compared with universities. The lowest level of coverage at MOE institutions is found at public general upper secondary schools, the highest at the Rajabahts (where over a third of the enrollment is receipt of a loan). We may note also the high level of coverage at private institutions, in comparison with coverage at parallel public institutions: about a third at private upper secondary schools and some 20 per cent at private universities. 3

13 d. A central problem with the extensive coverage of the scheme is that growth in the number of borrowers has not been matched by increasing loans budgets. It is the case that the value of total annual loan disbursements continues to increase (planned to exceed 24 billion Baht in 1999), but this is mainly to meet existing commitments to established borrowers. The loans budget for new borrowers is planned to remain fixed at six billion Baht, in nominal terms, at least until the year 2001; this implies a declining level of total support in real terms. A fixed loans budget for new loanees, together with a growing demand for student loans, has resulted in a drastic reduction in the average loan size for new recipients in The average loan size in 1997 was some 30 per cent higher in nominal terms than in 1998; measured in real terms (taking account of inflation) the gap would be even greater. This unplanned outcome militates against the objectives of aiding needy students and the principle of horizontal equity, to which we now turn. Figure 1 Student Loans Scheme Coverage, 1998 Rajabhat Up Sec: Private General & Vocational Up Sec: Public Vocational RIT University - Private Up Sec: Public General University- Public % Targeting Loans Budget Allocation a. The central purpose of the SLS, as outlined in the Act, is to offer financial assistance to poor students and to enlarge the educational opportunities of youngsters from low-income families to continue their education, up to the level of the bachelor degree. These objectives can be realized only through a well designed loans distribution system that targets loans to reach the client population. Unfortunately, the loans allocation mechanism adopted by the scheme is not designed to achieve this purpose. b. Rather than using a centralized loans clearing-house to service individual loans applications from potential borrowers, loans are distributed to students at the level of the educational institution, be it the school, college or university (Figure 2). Each institution receives a loans budget, through a top-down allocation mechanism that has been superimposed on the organizational and administrative 4

14 structure of the educational system. This allocation mechanism poorly serves the twin objectives of targeting needy students and horizontal equity. Figure 2 Student Loan Fund Allocation Student Loan Committee MOE (F irst Paymen t Accou nt Subcommittee) MUA (Secon d Payment Account Subcommittee) Central Loan Allocation Provincial Loan Allocation Non-MOE Ministries Bangkok Departments Provincial Loan Committees Universities Departments Faculties Educational Institutions Educational Institutions Educational Institutions Student Loan Applications Student Loan Applications Student Loan Applications Student Loan Applications c. The allocation of the loans budget by the Student Loans Scheme Committee (SLSC) to the Ministry of University Affairs (MUA) and the Ministry of Education (MOE) respectively, and the subsequent allocation of loans budgets down the line is largely arbitrary in terms of targeting needy students. The system works to ensure that, as it were, a fair share of the loans budget is received at each level of the ministry s organizational hierarchy, down to the level of the education institution itself, which actually distributes the loans to students. But, with the partial exception of MOE loans budget allocation to the provincial offices, this loans budget allocation system essentially ignores the scheme s central objective of targeting the poor. d. The MOE is considerably stricter than the MUA in setting criteria for loans distribution, in terms of the upper limit put on applicants household income, the maximum size of loan and the purposes for which loans may be received 5

15 (particularly in relation to accommodation). This leads to considerable inequalities in treatment, particularly of first degree students across the two ministries. e. Institutions are given an extensive degree of autonomy in the distribution of loans budgets, with the result that distribution policies differ widely from institution to institution. Some institutions focus on the very poor, while others distribute smaller loans to a larger number of recipients, including the less poor. Abuse is not uncommon at private institutions that have a vested interest in spreading loans widely, to cover tuition costs for a larger number of students; in the interests of augmenting enrollments. The result is that, in many cases, poorer students do not receive the loans to cover living expenses, that it is the intention of the SLS to provide. Such decentralized decision making, leading to differing outcomes for students of similar background and need, appears to be inappropriate for the distribution of loans; issues of horizontal equity and fair treatment are grossly neglected. f. A major strength of the present system lies in the successful targeting of the poor at the educational institutional level; loans are distributed in terms of need, with little queue jumping apparent. However, the system lacks horizontal equity because of the lack of objective criteria (in terms of targeting) in the allocation of loans budgets to institutions and the widely differing institutional loan distribution practices. Individual youngsters, with similar need, receive dissimilar treatment, even when enrolled in the same type institution Targeting the Poor The central objective of the student loans scheme is to provide loans to students from poor families, to enable them to continue study at the upper secondary and tertiary levels of education. How far the scheme succeeds in this objective depends on three main issues: reach, efficacy and equity. First, does the scheme indeed reach the target group? Which groups are not reached? Are some ineligible students in receipt of loans? Second, of those who are reached by the scheme and in receipt of a loan, does the loan indeed lead to continuation of study, thus preventing dropout? What would happen in the absence of a loan? Third, are loan recipients treated equitably? Do all potential recipients have equal chances of receiving a loan? Reach of the Scheme: the Egg Diagram a. The concept of targeting and reach may be clarified by reference to Figure 3. The outside (largest) egg relates to the total population of school age and includes both youngsters who have left the school system and those currently attending (the middle egg). Within the schooling population, some students are in receipt of loans (the smallest egg). On the basis of poverty measures, only part of this total youth population is defined as poor located in the north-east section of the chart. Total loan recipients are shown in sections A and D of the chart. A are poor and thus entitled to a loan, while D are not poor but are in receipt of loans either because the eligibility criteria used do not coincide well with poverty definitions or the loans scheme is being abused. The targeted school population consists of all poor youngsters enrolled in school (sections A and B); of these only 6

16 A receive loans and B are not reached. Areas E and F refer, respectively, to nonpoor students who do not receive loans and non-poor youth outside the schooling system. The total target population consists of A and B plus C (a proportion of poor youngsters that have left school but could be drawn back through the availability of loans). Monitoring, on a continuing basis, will show how successfully the targeted population is being reached. Figure 3 Reach of the Scheme Non-Poor C B Poor A D Non School Population Students : Non Loan Recepients Students : Loan Recepients A = Target Group: Loan recepient B = Target Group: Not receiving a loan C = Target Group: Potential Enrollees D = Not Target Group: Loan recepients b. The relative sizes of the different areas are likely to differ for upper secondary and tertiary education, respectively. Since the MUA, which is responsible for the large part of tertiary education, defines an over-generous eligibility ceiling (of up to 300,000 Baht and clearly above the poverty line), area D is likely to be relatively larger for tertiary education than for upper secondary education, which employs an upper income limit of 150,000 Baht. However, area C is likely to be relatively larger for upper secondary education, given current policies to increase significantly the upper secondary enrollment rate. c. Using data from the 1996 socioeconomic survey (SES) conducted by the National Statistical Office, we ran a special program to identify the total number of youngsters aged (the approximate age of upper secondary enrollment) in poor households. We defined a poor household in terms of the MOE loans eligibility ceiling of household income below 150,000 Baht. In 1998, there were 1,710, 000 poor youngsters (aged 15-17) in the population, 1,155,000 of whom 7

17 were attending school. Some 400,000 high school youngsters, or about 35 per cent of poor upper secondary school students, were in receipt of loans; the vast majority of needy students (65 per cent) were not. The proportion of poor students receiving a loan has grown each year with the increased take-up of the scheme (10 % in 1996 and 27 % in 1997). Of the total target population of poor youngsters (including poor youth who were out of school), loans are received by only about a quarter of the ultimate target population (up from 6% and 18 % in 1996 and 1997 respectively). d. However, there are differing intensities of poverty, related to the size of household income. Of the very poor (with household income of up to 50,000 Baht), over half a million were at school and more than a further quarter of a million were outside of the education system. Thus the total number of loans given in 1998 not only covered but a small proportion of the target population; it was in fact insufficient even to cover the half a million very poor students. Had all loans been targeted to reach only the very poor, 20 per cent of the very needy student would have gone lacking. The essential question is whether these loans reached the very poor or were taken up by students who were less in need? Unfortunately, due to the lack of a monitoring system, it is not possible to answer this question for upper secondary education. e. From the foregoing discussion, it appears that the SLS is to be seen less as targeting scheme than one for screening. The distinction between targeting and screening is subtle, but important. A loans scheme based on screening focuses on the vetting of applications for eligibility and inclusion, on the basis of defined entry criteria. It is less concerned with the actual composition of those included (and whether they constitute the group most in need of the loan), nor with those self-excluded, but eligible, individuals who do not apply for a loan. It is essentially passive in approach, emphasizing legal entitlement rather than satisfying need. A targeted scheme, on the other hand, actively focuses on those most in need; it aims at reaching out to a target population defined in terms of those most deserving of help, including such less easily reached targets populations as student dropouts and the rural poor. The challenge facing the Thai student loans scheme is to reorient its approach, based on improved information bases and monitoring, in the direction of a more highly targeted, focussed, proactive scheme Effect on Access and Dropout Prevention a. For those who are reached by the scheme and in receipt of a loan, does the loan indeed lead to continuation of study, thus preventing dropout? What would happen in the absence of a loan? These issues were probed in the special Loans Interview Survey carried out for this review. Loan recipients were asked: If there were no loan, how would you manage to continue your studies? Would you quit? The vast majority of loan recipients said that they would not drop out in the absence of a loan; in all, only 30 per cent said that they would do so. This finding is very suggestive. On the positive side, it does appear that loans are effective in preventing discontinuation of studies amongst a significant proportion of current borrowers, but they represent a minority. For the most part students claim that 8

18 they would be able to suffice by working, being more economical or by resorting to family for help. b. Respondents were also asked how their parents would react to financial difficulties: would a child be taken out of school and, if so, which gender? In twothirds of the cases, parents were not expected to withdraw their children from school. Of the remainder, there was no difference in the tendency to take a boy or a girl out of school. c. We have noted that the majority of eligible students are not in receipt of a loan. In the Loan Interview Survey, eligible students who have not borrowed were asked why they had not applied for a loan. Evidently, this was not because of a lack of awareness of the scheme: over 90 per cent were informed about the loans scheme, the vast majority receiving information from their educational institution. About half confirmed that they did not need a loan; a further quarter did not apply, either because of parental opposition or because they feared entering into debt. Some 20 per cent were interested in receiving a loan but did not do so for technical reasons: lacking the necessary documentation to apply, had applied but had not been accepted or had found the application form too complicated. Clearly, there remains some considerable scope for extending the reach of the scheme amongst eligible students; but appropriate steps must be taken to ease application procedures and explain more deliberately the benefits and obligations of borrowing. d. What of working youngsters? The central objective underlying the establishment of a targeted loans system is to aid needy students and encourage continuity of education and thus a larger enrollment of youngsters from poor families. With the subsequent onset of the economic downturn there has been a new emphasis on student loans as a means of preventing dropout from the educational system. However, there are two opposite tendencies at work here. On the one hand, conditions of higher unemployment and lower earnings possibilities, result in reduced opportunity costs of study and training. On the other hand, with lowered household income, there is increased pressure on all family members to contribute to family income and to be available to take up earnings opportunities that arise. Parents of out-of-school youngsters and potential college students may be loath to assume loan burdens in such conditions. A period of economic downturn may not be the most favorable for making long term education investments (unlike short training courses) nor taking on additional, future loan repayment commitments. Trust in a better future may not be manifest amongst the poor, who may be influenced more by the clear evidence of diminished employment opportunities and lowered earnings than by economic forecasts. Time horizons of the poor tend to be short and there is evidence that they have a tendency to discount the future more highly than higher socioeconomic groups e. These considerations may be more strongly applicable to upper secondary schooling than to education at the tertiary level. This is particularly so for general secondary education, whose main benefit stems from it serving as a bridge to higher education rather than earnings enhancement. These arguments also add support to our recommendation that loans for secondary school children be 9

19 replaced by a regime of targeted grants. Certainly, in terms of attacking the problem of dropout, grants are likely to be the more powerful instrument Horizontal Equity a. Inequalities in the system of loans distribution are to be expected and are justified if students have differing needs. Tuition fees at vocational schools are higher than at general secondary schools; students studying away from home may be in need of an accommodation loan while those living at home may not. These are examples of acceptable inequalities. Inequities arise when loan recipients of the same economic status and level of need receive loans of different size. b. Inequities across individuals may arise in two ways. They may be formally sanctioned by institutional arrangements for loan allocation and distribution. For example, the MOE and MUA operate on different criteria in relation to eligibility (household income level) and to size of loan maxima for undergraduate students. But inequities may be unintended, and arise because of a lack of sufficient information generation, monitoring and feedback. An example is loans distribution policies of individual education institutions. These institutions are free to fashion their own policies with regard to the size and purpose of individual loans, within the maxima set by the ministry. Declining budgets for new loans (in real terms) combined with a policy of renewal of loans for current borrowers at existing rates, have resulted in differences in treatment between students of similar need, even when enrolled at the same institution. c. In the Thai loans scheme these kinds of inequities are rife. There are inequities of treatment between similar institutions in different education ministries; inequities amongst institutions within the same ministry; and inequities in treatment of students enrolled at the same educational institution. However, is one particular area, that of gender, we did not find any evidence of inequity in treatment, neither in the Loans Interview Survey nor in the field visits. Indeed, in some institutions, because of a larger female enrollment, the percentage of total loans received by female students was in excess of that of males. At Thammasart University, female students constituted 63 per cent of all borrowers in 1997 (up from 59 per cent in 1996). 3.3 Finance a. A self-financing student loans scheme requires initially large and growing disbursements, as the scheme builds up. In principle, total disbursements should then level off as the first cohorts of borrowers complete their studies and exit the education system; subsequently the loans fund becomes self-sustaining as loan repayments increasingly finance the loans to new borrowers. b. The expectation that the Thai SLS will operate in this way seems to be widespread both in central government circles and at the educational institution level. In practice, such a positive outcome is highly unlikely. An extremely low (one per cent) interest rate, repayments in nominal (rather than real) terms; a lengthy repayment horizon of 17 years (including a two years grace period); 10

20 administrative costs and inevitable default: all these factors combine to produce a highly subsidized loans scheme. It was found, on the basis of computer simulations, that current repayment conditions imply that the typical borrower will return only about 20 per cent of the loan received, in real terms (implying the receipt of a hidden grant equivalent to some 80 per cent of the value of the loan), in real terms. Moreover, the expected burden of repayment is extremely light for the typical student. Annual repayments out of annual income will be in range of per cent only. These calculations point to an overly generous loans scheme, providing borrowers with unnecessarily highly subsidized loans, which may be repaid effortlessly out of higher income received on completing courses of education. c. The overall recovery ratio to the loans fund on typical loans will be even less favorable. Some students will not meet their repayment obligations and the administration of the loans fund is not costless. Taking account of the probability of repayment default and adopting a level of administration costs in line with that from international experience, results in recovery ratios in the range of 6 to10 per cent indicating that loans would be only marginally preferable to scholarships, on purely financial grounds. The broad coverage of the Thai scheme may result in economies of scale and lower administration costs; yet even so, recovery ratios exceeding 15 per cent are unlikely to be forthcoming. d. The loans scheme will not be fully self-financing, even over the long term; indeed, present arrangements imply a large and continuing call on the public purse. Given the clear social objectives of the SLS, a high level of subsidy may be justified; however, a continuing commitment of public funds at present levels is unlikely to be acceptable. If student numbers continue to grow and the demand for loans continues apace, this commitment will be increasingly onerous, particularly if the current level of support is maintained (in terms of the ratio of the number of borrowers to total student enrollments). e. It is government policy to encourage the growth of private universities as a means of increasing the number of university places, at low public cost. The average size of loan received by students at private institutions (particularly universities) exceed those for students enrolled at public institutions, because of considerable differences in tuition fee levels. The low level of expected repayments in real terms (in relation to the size of a loan) implies a 90 per cent subsidy of the costs of private education. The expansion of private education, with a growing commitment to provide highly subsidized loans, may not represent a cost-effective method of increasing university enrollments. 11

21 4. Towards Reform of the Student Loans Scheme a. The loans scheme, as it now stands, with its excessive level of subsidy, is not viable. Change can go in one of two directions. With such a high hidden grant element, a strong case can be made for replacing the present loans scheme by a regime of grants. In order to maintain the present level of net costs, the scholarship scheme would operate at a somewhat lower level of total budgetary support; there would be considerable savings in administration costs and, of course, the issue of repayment and default does not arise. Alternatively, steps could be taken to enhance the financial viability of the scheme by reducing the hidden grant element, through replacing the over-generous repayment conditions with more realistic ones that are somewhat closer to those of the market. This would imply a combination of higher interest rates, a shorter repayment horizon and an annual repayments correction for inflation. b. Our central recommendation for reform is a dual one. At the upper secondary level, the loans scheme would be replaced by a system of targeted scholarships. In parallel, the loans scheme would be retained for students at tertiary level institutions, but it would be radically reformed in the direction of improved targeting, the unification of loans conditions across all tertiary level institutions and improving the financial viability of the scheme through a hardening of repayment conditions 4.1 Organizational Reform: Improving Targeting and Horizontal Equity The major problem in developing a reform agenda is how to reform the present system of an arbitrary top-town allocation of loans budgets, that is combined with inequitable institution-based (decentralized) loans distribution. The direction advocated here is its replacement by a bottom-up system of loans applications to a central loans distribution body, which would both facilitate targeting and avoid horizontal inequity. Since the needs of secondary education differ dramatically from those at the tertiary level, reform will be considered separately for the two levels Tertiary-level Education a. The central elements of reform are: a.1. Replace the present system of top-down allocation of loans budgets to institutions and the institutional distribution of loans. a.2. Unify the present distribution system across the whole tertiary education sector (universities, RITs and Rajabahts) to provide parallel treatment for all students, whether enrolled in MUA or MOE institutions a.3. Distribute all loans from a central body to which applicants for loans would apply. The limited scope of the new Office of the SLS could be extended to include this function; it would be redesignated as the Students Loan Bureau (SLB) 12

22 a.4. Eligibility criteria would be the same for all applicants in terms of maximum parental income. There would be fixed schedules for the maximum size of loan available and its coverage of particular expenditure categories; all applicants are eligible for a maximum loan. However, it is recommended that eligibility criteria be revised, on the basis of additional research, to include additional elements such as family size and composition and number of children studying. a.5. Existing loan recipients and new borrowers would be treated on a more equal footing than at present. a.6. The actual cut-off parental income would vary somewhat from year to year, depending on the size of the loans budget and the number of loans applicants a.7. It would be possible to use supplementary criteria, in addition to parental income, to meet societal objectives such as encouraging studying to enter shortage fields. However, it is arguable that these objectives are best dealt with through more favorable repayments conditions rather than weakening the principle of needs-based loans. a.8. All applications received would be placed in a central computer base. Loans would be distributed in ascending order of parental income size, until the total available loans budget has been allocated. Thus the uniform cut-off level of parental income would ensure focussed loans targeting and horizontal equity a.9. Loans officers at individual institutions would continue with some of their present functions: advice to students on loans procedures and eligibility, vetting the application form for validity (including information on parental income) and the preparation of a listing of eligible applicants. They would act as an intermediary between the applicant and the Loans Bureau but they would not distribute loans. To streamline the system and to reduce the administrative burden at the center, it is suggested that loans officers would submit all loans applications in bulk to the Student Loans Bureau. a.10. One disadvantage of the new arrangements outlined in the preceding paragraph is that institutional loans officers may now have little incentive to ensure that the loans fund is properly used. No longer authorizing the distribution of loans, they may now become partisan advocates of potential borrowers enrolled at their institution, leading to possibilities of abuse and inequities. These dangers could be avoided by the adoption of an alternative arrangement (along the lines of current practice at the MUA); however, this alternative does not represent our central recommendation. a.11. Under this latter alternative, the SLB would allocate a loans budget to each educational institution; the institution would be responsible for the distribution of its loans allocation among student applicants. At the present 13

23 time no objective criteria for loans allocation are available. This alternative, then, would be feasible only to the extent that objective criteria for loan budget allocation to institutions became available. Information on the socioeconomic profile of the student body at each institution could serve this purpose. In principle, such information could be generated without difficulty. b. Figure 4 presents these reform alternatives in schematic forms. Path B is the favored method adopted in many countries: a direct line between the student applicant and the central loans organization. Once the application has been vetted and approved, it is passed on to the KTB, the banking organization that administers the loan. In path C, the education institution acts as intermediary between the applicant and the SLB: it receives loans applications, vets them for eligibility and submits them in bulk to the SLB. Path D is broadly parallel to current arrangements: each education institution is allocated a loans budget by the SLB, which is distributed to eligible applicants. It differs from present arrangements in three ways. First, all tertiary institutions are brought within a common framework. Second, the loans budgets are allocated to institutions on the basis of objective criteria related to student need and thus ensures that loans are targeted on poor students. Third, actual loan size and eligibility criteria are standardized across educational institutions Secondary Schools In the case of secondary schools we present two alternative scenarios for reform, both of which employ a parallel administrative mechanism. One scenario envisages substantial reform of the present loans mechanism at the secondary schooling level. However our central recommendation (Scenario 1) is the replacement of loans for secondary students with scholarships (grants). a. Scenario 1: Student grants This scenario envisages dismantling the system of student loans at the upper secondary level and its replacement by scholarships (grants) to needy students. A number of considerations underlie this radical recommendation. a.1. Few countries have loans for school students. Technically there is little difference if the student-borrower is at college or high school: in the latter case, however, there may be an ethical issue involved with parents entering underage youngsters into debt repayment obligations that extend up to a twenty-year time horizon. a.2. At present levels of subsidy, the loans scheme is not sustainable.the hidden grant element is so large that it almost amounts to a grants scheme. But overt grants would be far sharper in effect in influencing poor families to continue with their children s education and administrative costs would be lower. The alternative, of increasing the scheme s financial viability through stricter repayment conditions, would considerably reduce the incentive effects of enhancing education participation. 14

24 Figure 4 Alternative Scenarios for Loans Allocation Reform Students Loan Bureau MOE Provincial Offices Tertiary Level Institutions Upper Secondary Schools Student Loan Applications A B C D a.3. Ongoing plans to introduce free upper level secondary schooling for all considerably weakens the case for the continuation of student loans for secondary students (to meet living expenses only). b. Scenario 2: Reformed loans scheme b.1. The second, less radical alternative is a reformed student loans mechanism for secondary school students. In our view the case for targeted grants rather than loans at the secondary level is compelling. Nevertheless, we present this scenario in case there is some imperative to continue with the loans program for secondary schooling. As we do not recommend this approach, it is presented in synoptic form only. 15

25 b.2. The scheme might work in the following way: b.2.a. The Student Loans Bureau would be voted a considerably larger budget to accommodate secondary school loans. b.2.b. There are two alternative arrangements for scenario 2, shown in Figure 4. A system of direct student applications to, and loans distribution by, the SLB is shown in path B. b.2.c. The preferred, decentralized alternative is shown in path A. Loans budgets would be allocated to the 75 provincial loans offices; the SLB retains a loans reserve for second round allocations. Students apply, via their schools for loans. Schools vet the applications and submit them in bulk to the provincial offices. Loans budget would be allocated to provincial offices, on the basis of such objective criteria as provincial poverty levels and the number of poor teenagers in the province (both students and dropouts). Some of these data sets have been developed as part of the research underscoring the present study. b.2.d. Each provincial office distributes its loans budget to student applicants using an identical mechanism employed by the SLB for tertiary level student loans. All applications within a given province are placed in a central provincial computer base and loans would be distributed in ascending order of parental income size, until the total available provincial loans budget has been distributed. b.2.e. Inevitably, the actual cut-off parental income for loans recipients will differ from province to province. Non-distributed loans budgets in provinces that had reached the maximum parental income limit would be returned to the SLB. Unallocated loans budgets, together with the loans reserve, would be allocated to those provinces where there remained qualified loans applicants The objective would be to equalize the cut-off parental income across provinces. b.3. In the first year of the scheme, it may be necessary to withhold a fairly sizeable reserve and unallocated first round loans may be large in some provinces. But once the general parameters of the scheme have been identified in the first year, only marginal second round budget allocations should be necessary in subsequent years. 4.2 Finance Reform a. Low rates of cost recovery, considerably below 20 per cent indicate that the loans scheme is not viable in its present form. We have recommended, though not only for financial reasons that that the upper secondary element of the loans scheme be converted to one of grants. For tertiary education, there appears to be no alternative other than to raise recovery rates to more acceptable levels. 16