46206 Introducing a Student Loan Scheme in Romania A Discussion Paper THE WORLD BANK MINISTRY OF EDUCATION, RESEARCH AND YOUTH THE WORLD BANK Millennium Business Center 2-4 Armand Călinescu Street, sector 2, Bucharest, Romania Tel: (40-21) 201-03-11 Fax: (40-21) 201-03-38 ©2009 The International Bank for Reconstruction and Development / The World Bank 1818 H Street NW Washington DC 20433 Telephone: 202-473-1000 Internet: www.worldbank.org E-mail: feedback@worldbank.org All rights reserved Report N. 46206 This volume is a product of the staff of the International Bank for Reconstruction and Development / The World Bank. The findings, interpretations, and conclusions expressed in this volume do not necessarily reflect the views of the Executive Directors of The World Bank or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. Rights and Permissions The material in this publication is copyrighted. Copying and/or transmitting portions or all of this work without permission may be a violation of applicable law. The International Bank for Reconstruction and Development / The World Bank encourages dissemination of its work and will normally grant permission to reproduce portions of the work promptly. For permission to photocopy or reprint any part of this work, please send a request with complete information to the Copyright Clearance Center Inc., 222 Rosewood Drive, Danvers, MA 01923, USA; telephone: 978-750-8400; fax: 978-750-4470; Internet: www.copyright.com. All other queries on rights and licenses, including subsidiary rights, should be addressed to the Office of the Publisher, The World Bank, 1818 H Street NW, Washington, DC 20433, USA; fax: 202-522-2422; e-mail: pubrights@worldbank.org. ISBN 987-973-0-06089-8 CURRENCY AND EQUIVALENT UNITS Exchange Rate Effective October 21, 2008 Currency Unit = RON (New Romanian Lei) 1 Leu = US$ 0.3659 US$1 =Lei 2.7326 FISCAL YEAR January 1 – December 31 ACRONYMS AND ABBREVIATIONS CNFIS National Council for Higher Education Financing EU European Union EUROSTAT Statistical Office of the European Communities GDP Gross Domestic Product HE Higher Education ISCED International Standard Classification of Education MEF Ministry of Economy and Finance MERY Ministry of Education, Research and Youth OECD Organization for Economic Cooperation and Development PhD Doctor of Philosophy Degree RON New Romanian Lei UK United Kingdom UN United Nations USAMV University of Agronomics and Veterinary Medicine VET Vocational Educational Training Romania Minister of Education, Research and Youth: Anton Anton Secretary of State for Higher Education: Remus Pricopie World Bank Vice President: Shigeo Katsu Country Director: Orsalia Kalantzopoulos Country Manager: Benoit Blarel Sector Director: Tamar Manuelyan Atinc Sector Manager: Mamta Murthi Task Team Leader:. Lars Sondergaard TABLE OF CONTENTS Executive Summary ..............................................................................................................i I. Introduction ..........................................................................................................1 II. Some important changes in the past decade in higher education and background information on the sector ...............................................................................5 III. Challenges facing the higher education sector ................................................17 IV. What are the policy implications of the identified challenges? ...................... 28 V. Introducing a student loan schemes as an additional policy instrument in Romania ..............................................................................................................................29 VI. Estimating the fiscal cost of introducing a student loan scheme ...................38 VII. Conclusions and Recommendations .................................................................42 Annex 1: Enrollment rates in higher education ............................................................. 45 Annex 2: List of experts met during the paper preparation period ............................. 46 Annex 3: Examples of Current Student Loan Programs ...............................................47 List of Figures Figure 1: Enrollments in private and public universities .................................................................... 6 Figure 2: Total students in public institutions of higher education (full-time equivalent) ................. 7 Figure 3: World Bank estimates of full-time equivalent enrollment in public and private universities .......................................................................................................................................... 8 Figure 4: Students (ISCED 5-6) all ages - as % of 20-24 years old in the population ....................... 9 Figure 5: Public spending on higher education as a share of GDP ................................................... 15 Figure 6: Public expenditure on education as % of GDP (2005, except Romania (2008)) .............. 21 Figure 7: Estimate of distribution of support to students in public universities and the monthly amount of support ............................................................................................................................. 26 List of Tables Table 1: Educational attainment of 25-29 year olds (2005/06) ........................................................ 10 Table 2: Educational attainment of 19-21 year olds (2005/06) ........................................................ 11 Table 3: Total income of public universities (million RON) ............................................................ 13 Table 4: Four examples of funding by source from the perspective of public universities .............. 17 Table 5: Total spending on HE, by private/public and type of expenditure (2008) ......................... 22 Table 6: Estimates of tuition fees at private universities (annual fees in 2008) ............................... 22 Table 7: Estimate of private spending on living expenses, transportation and books (2008) ........... 23 Table 8: 2005-06 Direct Student Costs – (annual costs, based on 9-month academic year) ............ 24 Table 9: Estimate of the fiscal cost of different student loan scenarios ............................................ 40 ACKNOWLEDGEMENTS This discussion paper is the result of a joint effort undertaken by the World Bank and the Ministry of Education, Research and Youth of Romania. It was prepared by a World Bank team consisting of Bruce Johnstone (principal author, consultant), Pamela Marcucci (principal author, consultant) and Lars Sondergaard (principal author and task team leader, Economist), Mariana Moarcăş (Senior Operations Officer), Ana Maria Sandi (Consultant), and in close cooperation with Remus Pricopie (Secretary of State for Higher Education), Ion Ciucă (Director) and Zeno Reinhardt (Advisor to the Minister) from the Ministry of Education, Research and Youth, and Gabriela Jitaru (Head of Department) from the National Council for Higher Education Financing. The final version of the paper incorporates feedback received in three rounds of discussions with approximately 140 participants at workshops on student loans conducted in the spring and early summer of 2008. The team is grateful for extensive comments from Bank peer reviewers Maureen McLaughlin, Lead Education Specialist, Ronald Hood, Lead Economist, and Peter N. Materu, Senior Education Specialist, as well as for valuable feedback from Prof. Ecaterina Andronescu, Rector of the University Politehnica of Bucharest, Prof. Mircea Miclea, Head of the Presidential Commission for Education and Research, Prof. Gabriella Pasztor, Secretary of State for Minorities and Relationship with the Parliament, Ministry of Education, Research and Youth, and Prof. Adrian Miroiu, National School of Political Studies and Public Administration. The team gratefully acknowledges the support of Raluca Banioti and Svetlana Raykova who organized the team’s visits to Romania, the workshops and assisted in processing this report. Discussion Paper on Introducing a Student Loan Scheme in Romania Executive Summary Major changes The higher education sector in Romania has undergone major changes in the in the past past decade, both in terms of size and policies. Enrollments in public and decade have private universities have almost tripled since 1998, reaching close to 1 reshaped the million students in 2007/08. Policy changes include (i) shifts in the structure higher education of most study programs as a result of Romania’s participation in the Bologna sector in process; (ii) a reform of higher education finance (in 1999) involving a move Romania. away from input-based funding (i.e. money being linked to number of professors employed), towards enrollment-based funding; (iii) a substantial increase in overall funding in recent years; and (iv) since 2001, provisions allowing public universities to admit fee-paying students, which provide a major new revenue source for university expansion and quality improvement. Some slow-down However, as a result of a significant drop in the growth of the age-relevant in the pace of cohort, the brisk growth in demand for higher education has begun to slow change is (at least in public universities) and is likely to be more subdued in the providing an coming years. The pace of change is slowing, providing an opportunity to opportunity to take stock of the policies governing higher education, and make changes as take stock. needed. What are the As part of this stock-taking exercise, the Ministry of Education Research and challenges Youth (MERY) is considering the appropriateness of adding a student loan facing the sector scheme to its portfolio of policy instruments, and asked the World Bank to and are new prepare a discussion paper which would (i) identify the challenges facing policy financing higher education; (ii) articulate how a student loan program could instruments address some of these challenges; and (iii) examine the likely fiscal costs needed? associated with different types of student loan schemes. This paper seeks to contribute to the policy planning required for a new student loan program in Romania. Message 1: The The first main message of this discussion paper is: despite the impressive achievements achievements in higher education during the past decade, a number of have been challenges remain. These challenges are: impressive but (1) the high and continuously rising costs of higher education; challenges (2) the overall low state contribution to higher education; remain (3) the minimal level of cost-sharing and low fees paid by fee- paying students; (4) the high private costs of higher education; and (5) the current very uneven participation from a socio-economic perspective (see table below), which in addition to being a problem in itself, also implies that further increases in participation rates will involve tapping into a pool of students who come less prepared for higher education, both academically and in terms of financial resources at their disposal. i Equity is a serious problem in higher education: mostly wealthy and urban students participate Educational attainment of 25-29 year olds (2005/06) (%, self-reported degree obtained) Vocational Post- Middle / High secondary No Primary school Apprentice school or formal (grades (grades (grades 9- (grades foremen's Higher schooling 1-4) 5-8) 10) 9-12) school educ. Total Urban 1.1 2.0 7.7 15.4 39.6 7.0 27.2 100.0 Rural 3.0 5.7 34.7 24.6 25.9 2.3 3.7 100.0 Total 1.9 3.5 18.5 19.1 34.1 5.2 17.8 100.0 Incomes (by quintile) Poorest 6.3 12.2 39.6 19.2 19.6 1.0 2.0 100.0 Richest 0.2 0.3 3.6 10.0 34.5 8.2 43.1 100.0 Total 1.9 3.5 18.5 19.1 34.1 5.2 17.8 100.0 Source: World Bank calculations based on household survey data 2005/06. Message 2: The second main message is that the challenges place policy makers in a additional bind: substantially more resources will be needed for higher education and resources will be neither the state nor students and their parents seem capable of shouldering needed but it is the full increase. While there seems to be some scope for increasing public going to be spending, this will involve diverting tax revenue from other priorities difficult to mobilize and/or raising taxes, both of which are not particularly desirable. them without Moreover, while there seem to be wide-spread willingness of students and hurting equity even parents to contribute to the cost of higher education, total private costs further. (including living expenses) are already high relative to average incomes – prohibitively so for students from low income households. Message 3: adding The third main message is that introducing a student loan scheme would be a student loan an appropriate policy instrument to add but existing policy instruments scheme as an need tweaking as well. By introducing a student loan scheme, students additional policy could defer paying for their cost of education, and, as a result, be asked to instrument to contribute more than they are currently doing. Moreover, for the same address challenges amount of resources, a student loan scheme could provide a higher amount is advisable… of financial support to a higher number of poor students, providing more of them with the financial means to participate in higher education. Adjustments needed to existing policy instruments in higher are needed because focusing on student loans without addressing other issues of educational access and finance will not be effective in increasing participation or reducing inequities. …but existing It appears that current government financial support for students reinforces policy instruments the inequitable outcomes observed rather than ameliorating them. The need adjustments main reason for this is that almost all state support is guided by past as well. academic performance rather than an assessment of financial need. The figure below shows an estimate of how state support was distributed across the students in public universities (in 2008). It shows that approximately ii half were admitted as state-sponsored (based on their academic performance in high school) while the other half was fee-paying. It also shows that the bulk of the additional support was given to state-sponsored students on the basis of their stronger past academic performance. In fact, the only support mechanism which is not linked to past academic performance is the of “social scholarships programâ€? which, as the figure shows, is small both in terms of monthly support levels, and numbers of students. Since almost all state support is linked to a student’s past academic performance rather than an assessment of financial need, state support tends to reinforce rather than ameliorate the inequities observed in participation (Estimate of distribution of support to students in public universities and the monthly amount of support (in RON) (a negative amount implies that the student is paying, rather than receiving) (2008)) 1 Distribution of Per student 100% student population state support 8,133 Tuition waiver, and merit- based scholarship, living in student dorm 90% 7,063 Tuition waiver, and merit- based scholarship, no student (percent of total number of students in public HE institutions) 80% State-sponsored dorm 6,702 Tuition waiver and living in 70% student dorm 5,632 60% Tuition waiver, no student dorm, no merit-scholarship 50% 732 -338 Social scholarship, student -730 dorms, fee-paying 40% Social scholarship, no student 30% Fee-paying dorm, fee-paying 20% Living in student dorm, fee paying -1,800 10% Not supported at all 0% 2008 Source: World Bank estimates based on Eurydice data, and discussions with MERY officials, CNFIS, students and university representatives. Message 4: introduce The fourth main message is that – while politically more difficult in the student loans as part short term – international experience suggests that it is more effective to of a re-designed introduce a loan program as part of a comprehensive re-designed student student assistance assistance policy, rather than having to go back and fix inefficiencies package. (and fight the same political battles) later on. This is not to say that the introduction of certain types of loan programs cannot be a strategic first 1 The exact distribution does not exist because each faculty has some freedom in designing how it distributes the public resources ear-marked for scholarships. MERY is currently in the process of contacting all public universities to get a more precise figure. iii step in a longer-term plan, which it clearly can (and has been in many countries), but rather that it is important to locate the aims and operations of the loan scheme within the broader policy picture from the start. Recommendation 1: The paper recommends introducing a minimally subsidized loan scheme Introduce a loan that is integrated into the Romania’s student financial assistance policy. scheme but pay close In order to do this, however, the loan scheme would have to: attention to design ï‚· Be part of a comprehensive package of cost-sharing with loan issues to minimize amounts sufficient to cover tuition fees and living costs; fiscal costs ï‚· Have reasonable repayment expectations with repayment obligations spread over a long enough time to be manageable; ï‚· Have automatic deferral options that are activated in the event of unemployment, prolonged illness or loss of work, maternity, etc. ï‚· Be designed to minimize costs by keeping the base interest rate close to the government’s own borrowing cost, minimizing servicing and collection costs and minimizing defaults by assuring that the students view the loans as real obligations, by requiring some form of repayment plan to be agreed upon before graduation, and by engaging in good collection practices; ï‚· Be designed to share risk between government and cosignatories. Co-signatory requirements may need to be limited to moral persuasion and loss of credit reputation so as to avoid losing those parents who have insufficient assets to pledge. Recommendation 2: The paper also recommends redesigning government scholarship Redesign existing programs in such a way that the bulk of fiscal resources is spent on the policies governing financially weakest students, Some limited amounts should still, as student assistance to incentive measure, be allocated to the very best-performing students better target needy regardless of means. This approach will be more effective in encouraging students desired outcomes. Under the current financial support schemes, the bulk of fiscal resources end up supporting students who would most likely have pursued a tertiary degree even without the public support. Moreover, fiscal resources end up supporting students who are more likely to come from resource-strong families who have motivated them, paid for tutoring, and provided them with an environment conducive to learning. Next steps: issue Should policy makers decide to introduce a student loan scheme, the political decision; following next steps will be needed: assign staff; and 1. Issue a government decision to disburse the first student loans by, shift focus to say, October 2009. The decision should contain a broad estimate implementation of the additional fiscal resources needed in the first years of the arrangements program. 2. Assign staff to work full-time on designing student loan scheme, addressing each of the issues raised in sections V and VI of this paper (e.g. what is the loan scheme(s)’s primary objective, and what monitoring arrangements will be introduced to assess whether those objectives are reached? Who will be eligible to receive the loans? What will be the terms of the loans? What are the repayment arrangements? etc.) iv 3. Shift the focus beyond policy design questions to implementation arrangements, including what will be the role of the main agents in any student loan scheme (governments and ministries; public agencies; banks and other capital sources; universities and colleges, parents and other co-signatories; and collection and servicing agents) and how will the main functions (such as setting eligibility terms, originating loans, subsidizing loan, providing capital, and servicing and collecting loans) be carried out? v Discussion Paper on a Student Loan Scheme in Romania World Bank Technical Assistance on Higher Education Reform I. Introduction The higher education sector in Romania has undergone major changes in the past decade, both in terms of its size and its policy environment. In terms of its size, enrollments in public and private universities have almost tripled since 1998, reaching close to 1 million students in 2007/08. With public universities unable to meet the burgeoning demand for higher education, the number of private universities has grown rapidly (reaching around 50 institutions in 2007/08). While their numbers has remained roughly constant in the past decade, the size of each public university has expanded significantly, as reflected in an increased number of professors, and an increased number of faculties in each university. In terms of its policy environment, four policy changes have been particularly important in recent years: first, Romania’s participation in the Bologna process has required substantial legal and institutional changes and changed the structure of study programs, shifting the majority of students’ regular study period from a five-year degree to a two-phased three-year bachelor, followed, possibly, by a two-year master’s degree. Second, funding for higher education has undergone major reforms (in 1999), moving away from input-based funding (i.e. money being linked to number of professors employed) to increasingly being based on the number of student enrolled, and, although being implemented at a slower pace, incorporating elements of quality. Third, overall funding has increased substantially in recent years, rising from less than 0.5 percent of GDP in 2000-2005 to around 0.8 percent in 2007 and 2008. Fourth, public universities have been permitted to admit fee-paying students, providing them with a major new source of revenue to finance their expansions and quality improvements. The pace of change is slowing somewhat, providing an opportunity to take stock of the policies governing higher education, and make changes as needed. Reflecting a significant drop in the age- relevant cohort, the brisk growth in demand for higher education has already started to slow (at least in public universities) and is likely to be more subdued in the coming years, allowing policy makers a more temperate environment for stock taking.2 Adding to this more favorable environment is the fact that, after almost a decade of planning and work, most of the legal and institutional reforms suggested by the Bologna process have now been formally adopted.3 As part of this stock-taking exercise, MERY is considering the appropriateness of adding a student loan scheme to its portfolio of policy instruments available at its disposal. To this end, MERY asked the World Bank to prepare a discussion paper which would (i) identify the challenges facing financing higher education; (ii) articulate and clarify how a student loan program could address some of these challenges; and (iii) examine the likely fiscal costs associated with introducing different types of student loan schemes4. The term student loan is used throughout the paper to 2 The number of live births dropped dramatically in the years following the collapse of the CeauÅŸescu regime (from around 375,000 per year, on average, in the late 1980s to 315,000 in 1990 and 275,000 in 1991). This significantly smaller cohort of students born in 1991 is graduating from high school in or around the year 2009. Source: Eurostat 3 See Korka (2008) 4 The World Bank support to this task is through technical assistance grant in the area of higher education finance. The assistance is expected to help accelerate institutional reform, address fiscal vulnerabilities, and help design a student loan scheme that would contribute to reaching the country’s higher education objectives. The World Bank team is composed of team leader, Lars Sondergaard, Economist, Mariana Doina MoarcÄ?ÅŸ, World Bank Operations Officer and three independent consultants, Ana Maria Sandi, D. Bruce Johnstone and Pamela Marcucci. 1 indicate any obligation to repay a sum of money advanced to the student either for tuition or other fees or for some or all of the expenses of student living (i.e. food and lodging) to be repaid in the future.5 To ensure that this discussion paper would reflect the views of a broad group of stakeholders, MERY and the World Bank hosted three workshops in the period February to June 2008. These workshops were attended by Romanian stakeholders (university rectors, representatives of student unions, experts from MERY, Ministry of Economy and Finance (MEF) and representatives from the National Bank of Romania). During the workshops, international experts were invited to share information on the theory and practice of loan schemes in other parts of the world and to start wider discussions on: (i) the challenges facing higher education in Romania, focusing in particular on those related to its financing; (ii) possible objectives of a student loan scheme; and (iii) how such a scheme could be accompanied by other policy changes to introduce greater cost sharing in higher education. During the third workshop, the main messages of the draft discussion paper were presented and the feedback received from participants has been incorporated into this final version. The first main message of this discussion paper is: despite the impressive achievements in higher education during the past decade, a number of challenges remain. These challenges are: (1) the high and continuously rising costs of higher education; (2) the overall low state contribution to higher education; (3) the minimal level of cost-sharing and low fees paid by fee-paying students; (4) the high private costs of higher education; and (5) the current very uneven participation from a socio-economic perspective, in addition to being a problem in itself, implies that further increases in participation rates will involve tapping into a pool of students who come less prepared for higher education, both academically and in terms of financial resources at their disposal. The second main message is that, taken together, the challenges place policy makers in a bind: substantially more resources will be needed for higher education and neither the state nor students and their parents seem capable of shouldering the full increase. While there seems to be some scope for increasing public spending, this will involve diverting tax revenue from other priorities and/or raising taxes, both of which are not particularly desirable. Moreover, while there seem to be wide-spread acceptance of students and parents to contribute to the cost of higher education, total private costs (including living expenses) are already high relative to average incomes – and these private costs are prohibitively high for students from low income households. The third main message is that introducing a student loan scheme would be an appropriate policy instrument to add to the instruments currently available to address the bind posed by the challenges but existing policy instruments need tweaking as well. By introducing a student loan scheme as a new instrument, students could, in effect, be allowed to defer paying for their cost of education, and, as a result, be asked to contribute more than they are currently doing. Moreover, for the same amount of resources, a student loan scheme could provide a higher amount of financial support to a higher number of poor students, providing more of them with the financial means to participate in higher education. However, focusing on student loans without addressing other issues of educational access and finance will not be effective in increasing participation or reducing 5 Such an advance is a student loan whether the repayment obligation is a fixed schedule of repayments or is an obligation to repay a portion of future earnings or income (that is, whether the obligation is a conventional mortgage-type or an income contingent loan). Moreover, it is assumed that the obligation is properly called a loan whether the advance, or borrowed funds, pass through the hands of the student and are then paid to the institution as a tuition fee, or whether the funds are paid directly to the institution and the repayment obligation is incurred simply through matriculation or graduation or both. 2 inequities; existing policy instruments in higher education also need tweaking to address these challenges. In particular, it appears that current government financial support for students reinforces the inequitable outcomes observed rather than ameliorating them. The fourth main message is that – while politically more difficult in the short term – international experience suggests that it is more effective to introduce a loan program as part of a comprehensive re-designed student assistance policy, rather than having to go back and fix inefficiencies (and fight the same political battles) later on. This is not to say that the introduction of certain types of loan programs cannot be a strategic first step in a longer-term plan, which it clearly can (and has been in many countries), but rather that it is important to locate the aims and operations of the loan scheme within the broader policy picture from the start. The messages of this discussion paper resonate roughly with the priorities that emerged from the consultations via workshops and meetings with key stakeholders. First, the majority felt that a student loan program in Romania should be aimed at increasing participation in higher education by addressing the financial barriers that preclude access for the disadvantaged students, particularly those from poor rural areas. Second, they felt that a student loan program should support specific governmental policies such as increasing the number of students in specific high priority fields including science, technology, engineering and mathematics. There was also nascent recognition among the stakeholders that these aims can only be reached in a context of increasing resources for higher education and that student loans may also serve (especially for the fee paying students) as a vehicle for allowing students and their families to contribute to the costs of higher education. The discussion paper is intended to help policy makers and key stakeholders assess the merits of introducing a student loan scheme. The paper aims to achieve this by explaining what student loans area, how they work, and how they could address some of the challenges facing higher education in Romania. It is also hoped that any discussion of the merits of a student loan scheme will be helped by the paper’s presentation of facts and figures on the sector’s evolution over the past decade, and its assessment of the current challenges facing the sector. Moreover, the paper presents rough estimates of the likely fiscal costs associated with different types of loan schemes. Next steps: What the paper does not intent to provide, however, is to a step-by-step guidebook for how to implement a student loan scheme. For instance, the paper does not discuss staffing or training needs of a possible student loan agency, or provide legal advice on how to negotiate contracts with commercial banks, or provide detailed guidance on how to improve collection rates. Such detailed technical assistance is clearly what is needed in the next stage if a decision is taken to introduce a student loan scheme. In particular, should policy makers decide to introduce a student loan scheme, the following next steps will be needed: 1. Issue government decision to disburse first student loans by, say, October 2009. The decision should contain a broad estimate of the additional fiscal resources to be allocated for this purpose in the first years of the program. 2. Assign staff to work full-time on designing student loan scheme, addressing each of the issues raised in sections V and VI of this paper (e.g. what is the loan scheme(s)’s primary objective, and what monitoring arrangements will be introduced to assess whether those objectives are reached? Who will be eligible to receive the loans? What will be the terms of the loans? What are the repayment arrangements? etc.) 3. Shift focus beyond policy design questions to implementation arrangements, including what will be the role of the main agents in any student loan scheme (governments and ministries; public agencies; banks and other capital sources; universities and colleges, 3 parents and other co-signatories; and collection and servicing agents) and how will the main functions (such as setting eligibility terms and loan term parameters, originating loans, bearing risk of default, subsidizing loan, providing capital, and servicing and collecting loans) be carried out? The discussion paper is divided into seven main sections. Following this first introduction section, the second section discusses some important changes that have occurred in the past decade in higher education. The third section identifies a number of challenges facing the sector. The fourth section discusses what the policy implications of those challenges are. The fifth section introduces the notion of a student loan scheme and discusses how such scheme could address the identified challenges. The sixth section presents some rough estimates of the fiscal cost associated with introducing a student loan scheme, and the seventh section concludes. As a companion piece to the discussion paper, a Primer on Student Loans was written. It provides more detailed explanations of some of the concepts raised in the paper. Short descriptions of student loan schemes that are in operation in nine other countries are included in annex 3 to this paper.6 6 See Bruce Johnstone and Pamela Marcucci: Primer on Student Loans, mimeo, State University of New York at Buffalo. 4 II. Some important changes in the past decade in higher education and background information on the sector The higher education sector in Romania has undergone drastic changes in the past decade, both in terms of its size and its policy environment. Providing an encyclopedic overview of the changes that have taken is beyond the scope of this paper. However, selections of changes that are judged to be particularly drastic or important are discussed below. This section also provides some background on the higher education sector in Romania. II.a Enrollments in higher education have almost tripled in the past decade Enrollments in public and private universities have almost tripled since 1998, reaching close to 1 million students in 2007/08. There is some uncertainty surrounding the exact number of students enrolled in private university since not all private universities dutifully report their exact enrollments to the public authorities. While estimates for 2007/08 range widely – from 300,000 to 400,000 – the trend over the past decade is clear: demand for higher education has grown rapidly and public universities have been unable to meet this growing demand. To accommodate this demand, private universities have emerged, growing in numbers to around 50 institutions in 2007/08, and with enrollment in private universities accounting for approximately 40 percent of total enrollment in higher education in 2007/08 (see Figure 1).7 While the number of public universities has stayed roughly constant in the past decade (at around 55), each university has grown substantially in terms of its number of full-time professors, number of faculties, and number of enrolled students. This growth is all the more impressive because it occurred during a period of declining number of individuals in the regular university-age age group. The number of 18-24 year olds – the age group that still accounts for almost 70 percent of total enrollment in higher education – has dropped by almost 400,000 (or almost 15 percent) in the past decade. However, the proportion of this age group that decided to pursue a tertiary degree – i.e., the “participation rateâ€? – has increased substantially, from less than 11 percent in 1998 to almost 25 percent in 2006.8 Moreover, participation rates of older individuals have also expanded rapidly: almost 9 percent of 25-29 year olds were enrolled in higher education in 2006, compared to less than 3 percent a decade ago. As Annex table 1 shows participation rates have also grown for 30-34 year olds and 35-39 year olds. 7 The large increase in private enrollment in 2007/08 is, most likely, an artifact of merging two data sources (Eurostat for 1998-2006 and MERY 2007/08), rather than a real jump in enrollment in private universities. Most likely, historical data available in Eurostat underestimate enrollment in private universities since –as mentioned in the text – these institutions were not dutifully reporting their enrollments. 8 As a reference, around 55 percent of 19-21 year olds in Romania had completed a high school degree in 2005/06 (see Table 2). When calculating the proportion of upper secondary completers, both graduates from high schools (3 years of study) and VET graduates (2 years of study) should be included. 5 Figure 1: Enrollments in private and public universities, part-time and full-time, and full- time equivalent 1,200,000 1,000,000 800,000 (students enrolled) 600,000 400,000 200,000 0 2007e 1998 1999 2000 2001 2002 2003 2004 2005 2006 Full-time, public Full-time, private Part-time, public Part-time, private T otal in full-time unit Source: Eurostat (1998-2006) and MERY (2007/08) Consistent with the older profile of students and with the implementation of the Bologna structure of programs, enrollment in graduate programs has expanded more rapidly than enrollment in undergraduate programs. Detailed data for public universities from the National Council for Higher Education Financing (CNFIS, the council in charge of working out financing arrangements for public HE institutions) is roughly consistent with total enrollment data from Eurostat (shown in Figure 1) and show that most of the growth in enrollment – at least in public universities – has come from an increase in Master’s and PhD enrollments (266 percent and 537 percent increases respectively) as opposed to that of undergraduate students (10 percent increase). This increase largely reflects the move towards a HE structure consistent with the Bologna structure, where tertiary education students initially enroll as undergraduates and, only thereafter, apply for a graduate degree. This two-part structure differs from what was in place in the early 1990s in Romania, where most tertiary students enrolled directly in a five-year program. 6 Figure 2: Total students in public institutions of higher education (full-time equivalent) 600,000 500,000 (total students enrolled) 400,000 300,000 200,000 100,000 0 2003/2004 2004/2005 2005/2006 2006/2007 2007/2008 Total undergraduate students Total Master's students Total PhD Students Total students in HE (CNFIS figure) Total (Eurostat figure) Source: CNFIS and Eurostat The increases are also impressive because they were largely driven by an increased number of fee- paying students, both in private and in public universities. As already discussed and shown in Figure 1, enrollments in private universities have grown rapidly. However, enrollments in public universities were also largely driven by an increased number of fee-paying students (permitted since 2001 and discussed in more details below) as opposed to large increases in state-sponsored students (Figure 3).9 In 1998, more than two thirds of all tertiary students were state-sponsored. A decade later, seventy percent of all students were fee-paying. 9 The historical data (1998-2003) shown in Figure 3 are based on World Bank estimates (using Eurostat aggregate enrollment figures). 7 Figure 3: World Bank estimates of full-time equivalent enrollment in public (fee-paying and state-sponsored) and private universities (based on CNFIS and Eurostat figures)10 900,000 (enrolled students, part-time and full-time) 800,000 700,000 600,000 500,000 400,000 300,000 200,000 100,000 0 2007e 1998 1999 2000 2001 2002 2003 2004 2005 2006 Public, state-sponsored Public, fee-paying Private T otal, public and private Source: World Bank estimates based on CNFIS and Eurostat figures Despite the impressive increases, Romania’s participation rates continue to lag behind those in other EU countries. A commonly used indicator for international comparisons is total enrollment (of all ages) divided by the number of 20 to 24 year olds. In the case of Romania, in 2005, total enrollment (of all ages, in public and private universities) was around 740,000 and there were 1.6 million 20-24 year olds, resulting in a figure of 45.3.11 Figure 4 shows similar calculations for all EU countries in 1998, 2000 and 2005. The graph shows that Romania’s expansion in higher education occurred during a period when almost all EU members – and especially all new Member States – undertook similar expansions and, in some cases, even more impressive expansions. Using enrollment and completion rates for 2004, Coomans (2005) estimated that, in the EU, about 30 percent of the 25 – 34 year old age group will have achieved tertiary education by 2020 compared to below 15 percent for Romania. Although, enrollment rates have increased in recent years, the gap with the rest of the EU persists. 10 Enrollment in private universities is based on Eurostat figures (for 2001-2007). Figures for full-time equivalent enrollments for 1998, 1999 and 2000 are World Bank estimates based on full-time and part-time enrollments. Enrollments in public universities are using CNFIS figures for 2004-2007 and growth rates in Eurostat figures for 1998-2003 to estimate historical data consistent with CNFIS figures for 2004-2007. The split between fee-paying and state-sponsored is available for 2004-2007 (from CNFIS). For the years 1998- 2003, it is known that there were no fee-paying students prior to 2001. Moreover, it is assumed that the number of state-sponsored students stayed roughly constant at around 260,000 (full-time equivalent students) during the years 2001, 2002 and 2003. 11 Some care should be used when looking at this indicator because, as mentioned, the numerator contains enrollment of all ages whereas the denominator contains only the population age 20-24. Thus, a country with relatively more older (or very young) students will look better in this indicator, other things equal. 8 Figure 4: Students (ISCED 5-6) all ages - as % of 20-24 years old in the population 100 90 2005 2000 1998 80 70 (participation rate) 60 50 40 30 20 10 0 Czech Republic Belgium Luxembourg United Kingdom Romania Lithuania Bulgaria Slovakia EU27 Latvia Slovenia Finland Denmark Germany Greece Estonia Malta Italy Spain Portugal Poland Ireland Cyprus Netherlands France Sweden Hungary Austria Source: Eurostat One reason why Romania is still lagging behind other EU member countries in terms of participation rates is that students from rural and low-income households are only participating to a very small extent. Students from the wealthier and urban strata of the population in Romania are significantly more likely to go on to higher education. Almost half of the young people aged 24-29 from the richest income quintile (43 percent) have graduated from tertiary education compared to only 2 percent and 5 percent from the poorest two income quintiles (see Table 1). Similarly, about 27 percent of urban 25 to 29 year olds have completed tertiary education compared to only 4 percent of rural young people of that age. These structural inequalities are, in part, a function of the large proportion of the Romanian population (45 percent in 2005/06) living in rural areas, where the head of a household, on average, have completed fewer years of schooling, and have lower average incomes. Table 2 shows that, in terms of preparedness for higher education, only 37 percent of 19-21 year olds coming from rural households have completed a high school degree, compared to 68 percent of their peers in urban households. This is partly the result of there being very few upper secondary schools in rural areas, and therefore, almost all rural students who go on to upper secondary education, need to go to urban schools and are prevented to do so by the prohibitively high living costs (World Bank education note 2007). 9 Table 1: Educational attainment of 25-29 year olds (2005/06) (%, self-reported degree obtained) Vocational Post- Middle / High secondary No Primary school Apprentice school or formal (grades (grades (grades 9- (grades foremen's Higher schooling 1-4) 5-8) 10) 9-12) school educ Total Urban 1.1 2.0 7.7 15.4 39.6 7.0 27.2 100.0 Rural 3.0 5.7 34.7 24.6 25.9 2.3 3.7 100.0 Total 1.9 3.5 18.5 19.1 34.1 5.2 17.8 100.0 Incomes Quintile 1, poorest 6.3 12.2 39.6 19.2 19.6 1.0 2.0 100.0 Quintile 2 1.9 3.1 29.0 25.2 32.3 3.4 5.0 100.0 Quintile 3 1.1 1.6 17.1 22.6 42.5 5.2 9.8 100.0 Quintile 4 0.4 1.3 8.8 21.4 41.2 6.7 20.3 100.0 Quintile 5, richest 0.2 0.3 3.6 10.0 34.5 8.2 43.1 100.0 Total 1.9 3.5 18.5 19.1 34.1 5.2 17.8 100.0 Source: World Bank calculations based on household survey data 2005/06. There are, however, potentially a largely untapped pool of potential demand for higher education, hidden in poor households and households from rural areas: for instance, 37 percent of 19-21 year olds from rural areas have obtained a high school degree and, yet, five years later, less than 4 percent of all 25-29 years from rural areas have obtained a tertiary degree. Similarly, a very small proportion of the individuals from poorest households, which complete a high school, continue to pursue a tertiary degree. 10 Table 2: Educational attainment of 19-21 year olds (2005/06) (%, self-reported degree obtained) Vocational Post- Middle / High secondary No Primary school Apprentice school or formal (grades (grades (grades 9- (grades foremen's Higher schooling 1-4) 5-8) 10) 9-12) school educ Total Urban 0.8 1.8 12.8 16.1 67.6 0.5 0.4 100.0 Rural 2.5 4.9 32.5 23.2 36.6 0.2 0.1 100.0 Total 1.5 3.1 20.9 19.0 54.9 0.4 0.3 100.0 Incomes Quintile 1, poorest 4.2 8.2 36.5 22.5 28.1 0.3 0.2 100.0 Quintile 2 1.2 2.5 22.5 25.5 47.8 0.3 0.2 100.0 Quintile 3 0.7 1.6 16.1 18.8 62.2 0.4 0.2 100.0 Quintile 4 0.0 0.5 11.7 15.0 71.8 0.5 0.4 100.0 Quintile 5, richest 0.1 0.3 9.0 8.7 80.9 0.6 0.5 100.0 Total 1.5 3.1 20.9 19.0 54.9 0.4 0.3 100.0 Source: World Bank calculations based on household survey data 2005/06. There are reasons to suspect that the relatively low participation rate at tertiary level – especially of poor households – is partly a function of policies in place for the past decade. In particular, the very low participation rates in tertiary education of poor households is an outcome which is entirely consistent with the policy that provides financial support based almost entirely on merit rather than financial needs (see discussion below). While the private cost of higher education is only one of many factors determining whether an individual decides to pursue a tertiary degree, the high private cost in Romania may be an important tipping factor deterring relatively poorer households. As mentioned above, only thirty percent of all students enrolled in tertiary education are sponsored by the state and this selection is based almost entirely on their academic performance, rather than on a combined assessment of their performance and financial needs. The remaining seventy percent of total student enrolled in 2007/08 paid somewhere between 1,500-3,500 euros per year (including living expenses) to finance their education, with the cost varying to a large extent on whether they lived at home, in a student dormitory, or in a rented apartment (see Table 8). Some caution in interpreting the fast-growing enrollment figures is in order. As discussed above, the underlying demand for obtaining a tertiary degree appears very strong, especially considering that the growth is mainly from fee-paying students (whether in private or in public universities). However, discussions with university representatives, students and ministry officials suggest that the growing enrollments reflect a large number of students enrolled, but barely attending classes (and, according to some, nevertheless, obtaining a degree). To some extent this phenomenon is reflected in the growing proportion of part-time students (see Figure 1). What is not known, however, is the extent to which these part-time students eventually obtain degrees and what the quality of these graduates will be. In other words, access to education may be booming, but the quality of the graduates – if they all get that far – may have plummeted, at least in part of the system, because the system was not prepared to deliver quality education to a booming number of students. 11 II.b Major changes have been introduced to the policy environment The policy environment for higher education has changed substantially during the past decade. This section highlights four changes of particular significance: (1) the structural changes resulting from participation in the Bologna process; (2) the move from funding based on inputs to funding based on enrollment and quality; (3) the large increase in public funding to higher education; and (4) the introduction of fee-paying students in public universities. The first important policy change was Romania’s participation in the Bologna process. This has required substantial legal and institutional changes and changed the structure of study programs. Since the 2005/06 academic year, the higher education system in Romania has been re-organized from a two cycles system (5-6 years program plus PhD) into a three cycles system: Bachelor programs, Master’s programs and PhD programs compatible with the European qualification framework. Moreover, Government Decision 1175 of 2006 reduced the number of study fields to 15, and the academic year was broken down in two semesters, with each semester covering a minimum of 30 study credits. The total number of credits for a BA degree ranges between 180 and 240 credits. Currently, the main focus of the “Strategy for Development of Higher Education in the Period 2002-2010â€? – developed in line with the Bologna Process – has been compatibility assurance between the Romanian higher education system and other European systems. One of the concrete measures taken to achieve the objectives laid out in the strategy is the development of a new higher education legislative framework that is currently being debated. The second important policy change is that, in 1999, Romania moved from a system of input-based higher education funding to a mixture of per capita and quality-based financing. The new financing model is based on methodology developed by CNFIS, a council established as part of the reforms. The higher education budget is proposed by CNFIS to the MERY for the next fiscal year based on data provided by each university. As in the past, the budget discussion starts each year by the government approving the total number of study grants (see section on tuition and fees below) per study area and per university study cycle that it will fund based on the country’s economic and social development requirements. However, since the 1999 reforms were implemented, the MERY now distributes the study grants to universities also depending on quality indicators and institutional capacity through a Ministerial Order (rather than based on norms on how many professors each university is entitled to). Each university senate then distributes the study grants to study programs and study cycles according to its own strategy. The financing contracts between the government and individual public higher education institutions include the following components: core component and a complementary component. The nominal amounts disbursed in these groupings (as well as the size of public universities own sources of revenues) are reported in Table 3. Box 1 discusses core funding versus complementary funding in more details. 12 Table 3: Total income of public universities (million RON) 1999 2000 2001 2002 2003 2004 2005 2006 Core funding 175 270 380 477 633 847 1,041 1,215 Complementary funding 94 157 299 363 350 372 483 1,524 Own source revenue 109 167 236 385 489 618 723 878 Total income 378 595 915 1,226 1,472 1,837 2,248 3,617 Core funding 46% 45% 42% 39% 43% 46% 46% 34% Complementary funding 25% 26% 33% 30% 24% 20% 22% 42% Own source revenue 29% 28% 26% 31% 33% 34% 32% 24% Total income 100% 100% 100% 100% 100% 100% 100% 100% Source: CNFIS 13 Box 1: How are public universities receiving their funding from the state? Public universities receive their funding under two broad groupings: core funding vs. complementary funding. Below, each of these is explained in details. 1. Base (core) finance is used by universities to cover personnel expenditures (teaching and non-teaching staff payroll expenditures and travel expenditures) and material expenditures (for maintenance and administration, for functional materials and services, inventory goods, recurrent repairs, books and publications, and staff training, protocol, labor protection etc.). Seventy percent of the core finance is based on the number of physical students eligible by law for state budgeted financing (i.e. eligible for the study grant) as reported by the university in each form and field of education and calculated using a cost coefficient determined as the ratio of the financial effort required to train a student in a undergraduate program to the financial effort required to train an undergraduate student in economics, which is taken as the standard. The cost coefficient for social sciences and humanities, for example, is 1, while the cost coefficient for mathematics is 1.65 and for agronomy is 1.75. The student enrollment number is multiplied by the cost coefficient to get the unit equivalent student number which is then multiplied by the budget allocation for base finance as approved by the Budget law (at present, the per student budget allocation is € 600) less a reserve fund. The enrollment quota to be financed from state budgets for all education levels is established yearly through Decisions of the Government (MERY negotiates with the MEF). The placements (called “study grantsâ€?) financed from state budget are then allocated to public higher education institutions through a Ministerial Order. Thirty percent of the core finance is based on quality indicators that include teaching staff quality, teaching staff development potential, research work performance, research capacity use, quality of physical resources, quality of documentation, quality of academic, administrative and financial management and quality of social and administrative services provided to students. Each of these indicators is given a weight based on importance. The individual university allocations are computed separately using mathematical formulas involving the whole set of variables. 2. Complementary funding from the government is used by universities to cover among other things: i. Accommodation and food subsidies for students ii. Funds allocated on priority bases for capital expenditures, other investment expenditures and capital repair work; iii. University research funds allocated on competitive basis; iv. Social expenditures for students (scholarships, travel expenditures, etc.) distributed according to numbers of eligible students in each university; v. Procurement of materials and equipment Complementary funding is allocated to universities using formulas that are based on the numbers of eligible students. Source: CNFIS 14 The third important change worth highlighting is that – alongside an overall increase in spending on education at all levels – public spending on higher education has risen sharply in the past three years, driven mainly by increases in capital spending (Figure 5). After hovering at around 0.4 percent of GDP, total public spending on higher education rose to more than 0.8 percent of GDP in 2007 and 2008 (see Figure 5). The increase was driven by very large spending increases in MERY’s capital expenditure budget to finance university rehabilitation and new construction costs. Figure 5: Public spending on higher education as a share of GDP 0.9 Capital expenditure 0.8 Current expenditure 0.7 Total (share of GDP) 0.6 0.5 0.4 0.3 0.2 0.1 0.0 2002 2003 2004 2005 2006 2007 2008 Source: MERY and Eurostat (for GDP numbers) It is still unclear whether the recent boost in spending will be sustained over the medium-term and, to what extent, recurrent spending – kept so far largely unchanged as a share of GDP – will also be allowed to grow. While the recent increases in spending provided a much-needed infusion of additional resources to the sector, the additional increases came in the form of capital spending which are more easily reversible – should policy makers wish to use resources elsewhere – compared to recurrent spending. Some additional resources for higher education has been secured from the European Social Fund (756 million euro between 2009-2013) but this amount dwindles in comparison to total spending on higher education (i.e. assuming that the 756 million would be disbursed in equal installments over five years, EU funds would amount to only a 14.8 percent increase in total spending on higher education in 2008). Additional EU funds will be available on a competitive basis for infrastructure and research. The fourth and final change worth emphasizing is that, as of 2001, public universities have been permitted to admit fee-paying students, providing them with a growing source of revenue to finance their expansions and quality improvements. As Table 3 shows, public universities have had own sources of revenue even prior to the introduction of fee-paying students but these resources were stemming from sources with fairly limited growth potential. For instance, although state- sponsored students do not pay tuition, the institutions may collect fees from these students for taking longer to complete their programs than the time period limit set by law, for application and registration, and for the repetition of tests. With the introduction of fee-paying students in higher education, own source revenues grew rapidly, both in absolute numbers and in the overall 15 importance to public universities’ overall income. In 2006, the relative importance of own source revenues (as a proportion of total resources allocated to public universities) fell somewhat, reflecting the large infusion of additional capital investment resources (counted as part of “complementary fundingâ€? in Table 3). The figures in Table 3 are averages across all universities but they mask large variations between universities. To illustrate examples of these variations, Table 4 shows the sources of university funds as a percentage of total university earnings for four quite different public institutions. It is included to illustrate the vast differences in the relative importance of the sources depending on the type of university. The institutions reported in the table are: ï‚· the University Politehnica of Bucharest, a comprehensive technical university with over 26,000 students in 13 faculties; ï‚· the Bucharest Academy of Economic Studies, a large professional university with over 40,000 students seventy-five percent of which are graduate students; ï‚· the University of Petrosani, a regional technical university with only 7,000 students; and ï‚· the University of Agronomics and Veterinary Medicine Bucharest (USAMV – Bucuresti), an agricultural university with almost 15,000 students in 7 faculties. The different income structure largely reflects the ability of universities to sell their research and/or for the staff to participate in consultancies, and the extent to which they have fee-paying students. At one extreme, the University of Petrosani is most dependent on government funds with over 70 percent of its total budget coming from core (37 percent) and complementary government funding (33 percent). At the other extreme, just 42 percent of USAMV-Bucuresti’s total funding comes from the state. While the latter three receive similar proportions of their budget from government sources, there are large differences in the relative importance of the particular types of non- government funds that they earn reflecting differences in terms of their costs, missions and structure. The University Politehnica receives close to 40 percent and the USAMV 33 percent of their earnings from research activities, projects and consultancies (the USAMV receives an additional 12 percent from other economic activities) and less than 2 percent and 11 percent respectively from tuition fees, while the Academy receives only 8 percent of its earnings from research activities, but over 40 percent of its earnings from tuition fees. Close to 40 percent of its student body is fee paying and since its instructional costs are relatively low, tuition fees come close to covering total instructional costs. Only 7 percent of the students at the Politehnica are fee paying and the tuition fees charged (which exceed those of the Academy) still cover only 65 percent of the instructional costs. About half of the students at the USAMV are fee paying. 16 Table 4: Four examples of funding by source from the perspective of public universities (percentage of total earnings, based on budget for most recent year available) Polytechnic Bucharest University of University of University of Academy of Petrosani Agronomical Sciences Bucharest Economic (Budget, and Veterinary (2006) Studies (2007) 2008) Medicine Bucharest (Budget, 2008) Core funding from government 27.8 24.6 37.3 17.0 Earnings from tuition fees 1.7 42.4 17.2 11.0 Other earnings 3.5 0.0 2.2 12.0 Earnings from research activities, projects, consultancies 39.4 8.3 8.3 33.0 Complementary funding from government 23.8 29.3 32.7 25.0 Earnings from student residences and cafeterias 3.8 2.5 2.3 2.0 Total funding 100.0 100.0 100.0 100.0 Source: Information collected from web sites of the four universities. III. Challenges facing the higher education sector The pace of change is slowing somewhat, providing an opportunity to identify the key challenges facing the sector, and take stock of the policies governing higher education. This section identifies and discusses a number of challenges facing the sector. Some of the challenges facing the higher education sector are typical of higher education throughout Europe and other OECD countries; others more typical either of the transitional (i.e. post-Communist) countries of Central and Eastern Europe or of the less industrialized, lower per-capita GDP countries generally; and still others more a function of policies that Romania has chosen. The principal challenges are: 1. The high and continuously rising costs of higher education (in all countries); 2. The overall low state contribution to higher education (in Romania); 3. The minimal level of cost-sharing and low fees paid by fee-paying students (in Romania); 4. The high private costs of higher education (in most countries—even where tuition fees are low); 5. The current very uneven participation from a socio-economic perspective, in addition to being a problem in itself, implies that further increases in participation rates will involve tapping into a pool of students who come less prepared for higher education, both academically and in terms of financial resources at their disposal. 17 1. The high and rising costs of higher education Higher education in all countries tends to be expensive. The reasons are beyond the scope of this discussion paper, but are well known and generally accepted, having to do with the labor intensive nature of both teaching and research, the high (and thus expensive) technological content of both teaching and research, and the fact that higher education everywhere needs continuously to change — and therefore needs to add new programs and new faculty almost always faster than it can shed lower priority programs and less productive faculty. However, the more serious challenge is the tendency of higher educational costs to increase continuously over time, driven upwards by rising per-student, or unit, costs (generally unrelieved by increasing productivity as in the typical goods- producing sectors of the economy), and further driven upwards by rising enrollments (see Box 2 for an attempt to quantify some of these pressures). Enrollments, in turn, are driven upwards by a combination of increasing numbers of university-age youth (not the case in Romania) and further accelerated by rising participation, or the increasing proportions of secondary school graduates seeking to go on to higher education. The implications to the financing of higher education in Romania is that the revenue needs of higher education must also increase annually at rates considerably greater than the prevailing rates of inflation—and very likely (aside from some possible years of budget increases to “catch upâ€?) at rates greater than the governmental revenue is likely to increase. Box 2: Cost pressures facing higher education This box highlights a number of cost pressures facing higher education in Romania and attempts to quantify the size of such pressures. The analysis presented focuses on estimating what the likely rise in total costs are, irrespective of whether this cost will be born by the public or private sector. Several factors will place substantial upward pressures on costs in the coming decade, probably forcing total spending as a share of GDP to rise to 1.3-1.5 percent of GDP (from its current level of less than 1 percent): more demand for higher education, and more demand for higher quality education. While upward pressures could partly be absorbed through improving the efficiency of spending (e.g. accepting larger class sizes), the size of the upward pressures seem likely to require additional resources for universities. It should be noted that estimating the combined upward pressure on costs is difficult for a number of reasons, including because spending to upgrade the quality of HE could result in reduced demand for HE if students – as opposed to the tax payers – are asked to shoulder the increases without having access to a student loan scheme. This box addresses this complication by presenting ranges for likely cost increases, rather than a single figure. Below, some of the factors which will put upward pressures on costs are discussed in turn. The most important cost pressure stems from the likely higher demand for HE which could rise by another 30 percent in the coming decade – on top of the tripling seen in the past decade. This estimate is based on the fact that participation rates of Romanians in HE are still relatively low compared to their EU neighbors and that these gaps will gradually narrow. Specially, the estimated 30 percent increase assumes that participation rates will gradually converge (by 2015) to the current rates of the best-performing EU members (Finland). If so, total enrollment will more than off-set the projected decline in population numbers (i.e. the number of young adult in the age group 18-25 will decline by more than 1 million between 2005 and 2020). In fact, as box figure below shows, total enrollments in HE could rise to 1.3 million by 2015 and – driven by underlying trends in population numbers – only gradually revert to its 2008 level (of around 1 million) thereafter. These calculations are not factoring in changes in the movement of students in and out of Romania. While movements in both directions are likely to increase in coming years, forecasting the net change is difficult. Therefore, as a baseline, the implicit assumption in the figures presented in this box is that the net impact on student numbers will be zero. 18 The extent to which the likely higher demand for higher education will translate into higher actual enrollments will, partly, depend on the extent to which the government will increase the number of students it supports, or whether students will have to pay for their education themselves. Scenario: If actual enrollments were to rise to 1.3 million students and average total spending per student remain the same, this would cost an additional 0.2-0.6 percentage points of GDP, depending on whether the additional students would be attending public, private or a combination of both.12 Box figure 1: Projected demand for higher education 2008-203013 (total students in higher education, public and 1,400,000 1,200,000 1,000,000 800,000 private) 600,000 400,000 200,000 0 2008 2010 2015 2020 2025 2030 Using Finland (2005) participation rates Using Romania (2005) + increase in participation rates (by 3 percentage points) Source: World Bank projections using UN population projections, and under two different assumptions regarding growth in participation rates: (1) convergence to those of Finland (in 2005) by 2015; (2) 3 percentage points increase in Romania’s current participation rates. Another important upward pressure on costs is the fact that average spending per student is still substantially below those in the OECD, especially in private universities (see box figure 2 next page). With high likelihood, this gap will gradually narrow during the next decades as students and Romania’s accreditation agency continue to demand a higher quality of education, especially from private universities. Scenario: If the cost per student in private universities were to rise so that the spending gap to public universities was halved – currently, public universities cost four times more per student than private universities (7,172 RON per student compared to 1,800) – this would cost an additional 0.15 percentage points of GDP. If public universities were to narrow their gap with the rest of the EU and the OECD, total spending on higher education could easily rise 1.3-1.5 percent of GDP (from the current level of 1 percent of GDP). Other cost pressures which are more difficult to quantify will arise from: -Increasing costs of faculty in new academic fields (e.g., computer science, technology, biomedical science, finance, and the like) that compete most with private sector alternatives; - Increasing costs for quality improvements, including academic and administrative technology, scientific equipment, computing, and library acquisitions (whether on-line or hard copy); 12 As will be discussed below, total spending per student is significantly higher in public compared to private universities. 13 The figure for 2008 is an estimate of total enrollment in public and private universities in Romania (924,000). For 2005, Eurostat has total enrollment (then, estimated at 738,000) by age group. We combine these enrollment figures with UN population figures (also by age group) to estimate the participation rates for 2008. For the projections, we assume that each of the age groups converge to the participation rates in Finland (using data for Finland for 2005). 19 - Increasing capital costs, both for needed new facilities in locations outside the major metropolitan centers and for deferred maintenance on the existing physical plants; - Increasing costs of financial assistance for the increasing proportions that will be entering from low-income and/or rural backgrounds). Looking across the different sources of cost pressures and scenarios, cost pressures are likely to push up total spending on higher education by at least 0.3 percentage points of GDP and possibly by as much as 0.5 percent during the coming decade (from its current level of 1 percent of GDP to 1.3-1.5 percent of GDP). The question for policy makers is how much of these pressures can be met through efficiency improvements, and whether the remaining increases will be paid for through the budget or through increased cost-sharing: further increases in enrollments in private universities and/or allowing more fee- paying students into public universities. Box figure 2: Cost per student in OECD (2005), EU19 (2005) and Romania's public and private 14 universities (2008) 9,000 (US$, PPP, constant 2005 prices) 8,000 7,000 6,000 5,000 4,000 3,000 2,000 1,000 0 OECD average EU19 average (US$, Romania, public Romania, private (US$, PPP, constant PPP, constant 2005 university (US$, university (US$, 2005 prices) prices) PPP, constant 2005 PPP, constant 2005 prices) prices) Source: Davidson (2007), IMF WEO data base (for PPP adjustment factors), MERY and World Bank estimates 14 Davidson (2007) provides unit cost estimates for tertiary institutions in OECD and EU19 for 2005. The estimate of the unit cost in public universities in Romania was obtained by dividing total public spending by the number of students in public universities, and converting the figure into constant US$ PPP terms (using an adjustment factor calculated from IMF WEO data base). The estimate for private universities was calculated using an estimate of total costs of running private universities (estimated as described in the box on total spending in HE) and dividing total cost by the number of students in private universities (and converting the number into US$ PPP terms). 20 2. Low overall state contribution to higher education Notwithstanding recent increases, public spending on higher education in Romania is low by international standards. Figure 6 shows public expenditure on education and on tertiary education as a percentage of GDP for a number of EU and other selected countries. Despite several recent governments’ commitments to raising total public expenditure at all levels of education to 6 percent of GDP, this target was reached only in 2007. Moreover, at current levels of spending on higher education (at about 0.8 percent of GDP), Romania spends less than half of what the EU recommends for its members to achieve the Lisbon objectives. Most likely, years of under- financing in higher education in Romania will require years of additional financing simply to catch up. Figure 6: Public expenditure on education as % of GDP (2005, except Romania (2008)) 3.0 2.5 2.0 (percent of GDP) 1.5 1.0 0.5 0.0 Romania (2008) Slovak Republic Czech Republic Belgium Brazil United Kingdom Chile EU19 average Mexico Switzerland Slovenia OECD average Denmark Finland Iceland2 Germany Greece Estonia Portugal Italy Poland Turkey Ireland Spain New Zealand Netherlands United States Russia France Korea Australia Israel Norway Sweden Japan Hungary Austria Source: OECD and MERY 3. The minimal level of cost-sharing and the low rate of fees paid by fee-paying students Romania is behind most—but not all—countries in the extent to which students contribute to the cost of higher education (although the privately borne expenses, even for the governmentally- supported students, are still considerable as outlined below). While the current fee paying students bring additional revenue to the public higher education sector, the proportion of the per student total instruction cost that these fees cover differs from university to university and in no case do they cover 100 percent of the instructional costs at public universities. At the Politehnica, for example, the highest tuition fee is 2,800 RON/year, which only covers 65 percent of instructional costs, while at the Bucharest Academy of Economic Studies, undergraduate fee paying students pay 2,400 RON/year, which is very close to the total instructional cost (2,510 RON/year). Therefore, in all public universities, the state is making up for these shortfalls; and these uncovered expenses necessarily impact other competing priorities such institutional investments in quality and in institutional financial assistance for students. In fact, despite the high and growing number of fee-paying students, the private contribution to institutions of higher education – both private and public – covers only to 20 percent of total spending of these institutions. Romania spent approximately 1.0 percent of GDP on higher education institutions in 2008, the bulk of which were financed directly by public resources (see 21 Table 5). This estimate excludes private spending on living expenses but does include public spending channeled through educational institutions but earmarked to support students’ living expenses. Eighty percent of total spending on HE comes directly from the budget and is used to finance the daily costs of running around 50 public universities where more than half of the country’s students in HE are enrolled, and the remaining resources (20 percent) are raised through students fees (to both private and public universities). Table 5: Total spending on HE, by private/public and type of expenditure (2008) (million RON and as a percent of GDP) 2008 (% of GDP) Public expenditure on HE, total (ml) 3,762 0.8% of which: capital spending 1,218 0.3% of which: recurrent 2,544 0.5% of which: on student scholarships (not tuition waiver), transport, accommodation, etc 387 0.1% of which: to universities 2,157 0.5% Private expenditure on HE, total 1,431 0.3% of which: tuition fees to public institutions (2006 figure) 711 0.2% of which: tuition fees to private institutions 720 0.2% Total spending on HE 5,193 1.1% Nominal GDP (million) 470,075 Source: World Bank estimates based on MERY data, discussions with student representatives Although nearly half of total HE students are enrolled in private universities, the cost of operating these private universities appears to be only a fraction of the total cost of running the public universities. Specifically, Romania’s 55 public universities cost approximately 4.2 billion – 3.76 billion obtained from the budget and the remaining raised through student fees – to operate in 2008 (providing education to 524 thousand students) while the 67 private universities spent an estimated 734 million educating 400 thousand students.15 Spending by private universities was estimated by multiplying the number of students enrolled in private university by an estimate of the average tuition paid, and, thus, excludes other sources of revenue that could finance private universities (e.g. private donations, consultancy services provided by faculty at private universities etc). The estimate of the average tuition paid by students in private universities (see Table 6) is an estimate provided by MERY. Table 6: Estimates of tuition fees at private universities (annual fees in 2008) Social sciences 1,200-1,500 RON Math/physics 1,600-1,800 RON Medicine 2,100- 2,600 RON Simple average of fees 1800 RON Source: MERY estimate 15 Documenting the main reasons why private universities have substantially lower per student costs than public universities are beyond the scope of this paper. However, there are several plausible reasons, including (1) they are focusing on cheaper programs such as law and economics (as opposed to, say, engineering and medicine); (2) they produce less research; (3) they have larger class sizes; and (4) they have more part-time and long-distance learning students. 22 4. The high private costs of higher education Although the amount of revenue raised from student fees is relatively small (as discussed above in the third challenge), total private costs– including living expenses in addition to fees – faced by students in Romania are high relative to GDP per capita. In fact, the average annual total cost faced by a students and their parents could range from 10-30 percent of GDP (as shown in Table 8 below), with total private spending on living expenses amounting to 1.2 percent of GDP in 2008 (i.e. more than total public spending on higher education). These costs include transportation costs (the part which is not subsidies and included in the estimate above), costs of books, and living expenses. The 1.2 percent GDP has been estimated using enrollment figures and estimates regarding the proportion of the student population living at home, in subsidized dormitories, and in non-subsidized dormitories, as well as information on students’ expenses obtained through discussions with student representatives. Table 7: Estimate of private spending on living expenses, transportation and books (2008), million RON and share of GDP (% of 2008 GDP) Total private spending on living expenses 5,496 1.2% of which: living expenses for students in public universities 3,118 0.7% of which: living expenses for students in private universities 2,378 0.5% Source: World Bank estimate In a country with an average monthly income of € 500, the costs that are the responsibility of students and their families are considerable.16 Anecdotal evidence from some of the university officials interviewed indicates the difficulty that groups of students have in paying their fees and the overwhelming attention paid to costs by prospective students. Based on discussions with student representatives, the authors of this paper compiled a rough estimate of the costs born by student, presented in Table 8. The table shows the typical annual costs of different types of students: state-sponsored students (columns one, two and three), fee-paying students (columns four, five and six) and students attending private universities (columns seven and eight). The lodging and food costs are estimates for the thirty percent of students who live in student residences and the 31 percent who live in their own apartment, sublet or private flat (Eurostudent 2008). Lodging and food costs for the 38 percent of students who are living at home with their families are significantly lower for all categories of students (Eurostudent data). Those who live in the student residences pay approximately 198 RON (€ 56) per month, while those living in a rented apartment must pay at least 438 RON (€ 124) per month (for a shared apartment in Bucharest). 16 Calculated as nominal GDP (estimated at 470 bn RON in 2008) divided by population (using 2006 figure of 21.6 million), converted into euros at 3.67 RON/EUR and into a monthly income by dividing by 12. 23 Table 8: 2005-06 Direct Student Costs – (annual costs, based on 9-month academic year) Gov’t- Gov’t- Gov’t- Fee Fee Fee Private Private sponsored sponsored sponsored paying paying paying Students Students students students students students students students (in (living at (in student (in private (living at (in (in (living at private home) housing) apartment) home) student private home) apart housing) apart ment) ment) Application fee €50 €50 €50 €50 €50 €50 €25 €25 Registration fees €30 €30 €30 €30 €30 €30 0 0 Tuition fees 0 0 0 €500 €500 €500 €300 - €300 - €600 €600 Books/ €210 €210 €210 €210 €210 €210 €210 €210 equipment Lodging €405 €1710 0 €450 €1710 0 €1,710 0 Food €490 €520 €150 €520 €520 €150 €520 €150 Transportation €150 €150 €200 €150 €150 €200 €150 €200 17 Other personal €306 €306 €306 €306 €306 €306 €306 €306 expenses Total €1,640 €2,976 €895 €2,216 €3,476 €1,446 €3,221 - €1,191- 3,521 1,491 Source: Discussions with students; Consultation with university websites; Eurostudent findings using 2005/06 exchange rate of 1€ = 3.5258 RON 5. The current very uneven participation from a socio-economic perspective, in addition to being a problem in itself, implies that further increases in participation rates will involve tapping into a pool of students who come less prepared for higher education, both academically and in terms of financial resources at their disposal. Further expansions in enrollment figures will, inevitably, involve trying to get students who are less academically prepared and have fewer financial resources available to participate. The reasons are simple: (i) the much smaller cohort of individuals born immediately after the collapse of the CeauÅŸescu regime when abortion was legalized will graduate from high school this summer and next (the number of individuals born in 1990, 1991 and 1992 was 15, 26 and 30 percent smaller, respectively, than the number of individuals born in 1989); and (ii) as discussed, equity in higher education is a problem: the current profile of participants in higher education suggests that participation in higher education is very uneven, with participation dominated by students from wealthier and, predominantly, from urban households. While there is some scope to further increase participation rates of students from urban and relatively wealthier households, large further increases of the magnitudes seen in other new EU member countries will not be possible without broadening the pool of participants. Getting more of these students to enroll is not only a matter of money, but easing the financial burden will help. As mentioned, the private cost of higher education is only one of many factors explaining why a student enrolls. Thus, targeting additional funds to poor students, alone, will not be enough. But it will be a start. To finance such support, it seems timely to re-think how state support to student is currently allocated since current state support for students – which comes in different forms – is almost entirely guided by merit, rather than financial need. Romania’s support for students take different 17 Local transportation is subsidized: student buys half price ticket (not part of education budget) and (in some cases if fee paying) can get the other half reimbursed by the university. 24 forms: (i) whether or not to accept the student in a public university (resulting – even if the student is a fee-paying student – in a fiscal cost); (ii) if accepted in a public university, whether or not to accept the student as a state-sponsored or a fee-paying student; (iii) whether or not to offer the student a place in a subsidized student dormitory; and (iv) whether or not to offer the student one of four types of state scholarships (see Box 3 for a further discussion of each of these systems of support). Figure 7 shows an estimate of how state support was distributed across the students in public universities (in 2008). It shows that approximately half were admitted as state-sponsored (based on their academic performance in high school) while the other half were fee-paying. It also shows that the bulk of the additional support provided to students were given to students admitted as state-sponsored who were, again, admitted because of their stronger past academic performance. In fact, the only support mechanism, which is not linked to past academic performance is the tiny program of “social scholarshipsâ€?. However, as the figure shows, the program is small both in terms of its monthly support, and in terms of the number of students assisted. 25 Figure 7: Estimate of distribution of support to students in public universities and the monthly amount of support (in RON) (a negative amount implies that the student is paying, rather than receiving) (2008)18 Distribution of Per student 100% student population state support 8,133 Tuition waiver, and merit- based scholarship, living in student dorm 90% 7,063 Tuition waiver, and merit- based scholarship, no student (percent of total number of students in public HE institutions) 80% State-sponsored dorm 6,702 Tuition waiver and living in 70% student dorm 5,632 60% Tuition waiver, no student dorm, no merit-scholarship 50% 732 -338 Social scholarship, student -730 dorms, fee-paying 40% Social scholarship, no student 30% Fee-paying dorm, fee-paying 20% Living in student dorm, fee paying -1,800 10% Not supported at all 0% 2008 Source: World Bank estimates based on Eurydice data, and discussions with MERY officials, CNFIS, students and university representatives. As such, the bulk of fiscal resources end up supporting students who would most likely have pursued a tertiary degree even without the public support. Moreover, fiscal resources end up supporting students who are more likely to come from resource-strong families who have motivated them, paid for tutoring, and provided them with an environment conducive for learning. An alternative way to use fiscal resources is to allocate them where they are likely to matter the most, and/or encourage desired outcomes that, without the fiscal support, would otherwise not have happened. Doing so would imply using the bulk of state support on the financially weakest students, and only spending a limited amount on the resourceful students (e.g. as an incentive for the very best-performing students). 18 The exact distribution does not exist because each faculty has some freedom in designing how it distributes the public resources ear-marked for scholarships. MERY is currently in the process of contacting all public universities to get a more precise figure. 26 Box 3: Types of state support for students This box describes the different types of state support to students. Waiving tuition fees Based on academic performance in high school and/or on the results of the admission exams, the state pays (or waives) the tuition fees of approximately half of the students enrolled in public universities. Accommodation State funded students (and some fee paying students – see below) receive subsidized accommodation in the university residences and only pay a range of 40 to 80 RON per month. They also may eat in the subsidized cafeterias that are run by the universities. Approximately 30 percent of all students live in dormitories, but the proportion differs substantially from university to university. Some universities accommodate 80 percent of their students, while others only accommodate 20 percent. First and second year fee paying students who perform very well may qualify for state financing the following year and state funded students who do not perform well on exams may be dropped. The main movement between fee paying and state financed status occurs after the first year of university 19. State Scholarships Approximately one-quarter of all public higher education students (state sponsored and fee paying) receive scholarships20 financed by the state budget. Of these a minimum of 20 percent receive social aid scholarships targeted at needy students, 25 percent receive merit scholarships given to students with the highest marks, 30 to 40 percent receive study scholarships awarded to students with high academic achievements and 5 percent receive performance scholarships given to a very small number of the best students. Universities also award social allowances for needy students from university funds. Decisions of the MERY set the general eligibility criteria for the scholarships and the University Senates identify specific criteria. All government scholarships are indexed each year to inflation and are supposed to cover total student accommodation and meal costs. The scholarships do not cover the tuition fees of fee paying students, which may serve as a barrier to their being able to go on to higher education. Most scholarships are offered as part of the student’s entry package to an institution with the exception of social scholarships. These scholarships are means-tested and students must apply for them separately supplying income statements from their parents or certificates of retirement or unemployment or disability as well as statements from their local authority verifying the number of members in their family and other information. Applicants must have a certain score on exam that differs from year to year. The two types of social scholarships that are covered by the MERY complementary funding include: 1. Common scholarship for students whose family members earn below a certain minimum salary. 2. Scholarships for rural students (€ 100/month) who agree to work in rural areas for the same number of years as the number of years of their study. There are four types of merit based scholarships funded by the MERY complementary funding: 1. Merit scholarships: given to students with the highest marks. 2. Performance scholarships: given to a very small number of the best students (generally one per faculty). 3. Study scholarships: awarded to students with high academic achievements. 4. Olympic merit scholarships: based on international (academic) Olympic competitions. All students receive free medical and psychological care and pay only half price rates for public transportation. If there are left over places in the student dormitories once all of the state sponsored students have been accommodated, the left over spaces may be allocated to fee paying students on a merit basis. In some institutions, a certain number of places are reserved for needy students. Generally, the fees charged to state sponsored and fee-paying students are different. 19 Two universities in Bucharest report that there is a movement of 10 to 20 percent each year between fee paying students and state funded students. 20 They are referred to as scholarships (burse), but are really grants that cover living costs. 27 IV. What are the policy implications of the identified challenges? Taken together, the five challenges place policy makers in a bind: the first challenge highlights that more resources are needed, while the other challenges raise concerns about how to finance such increases. Clearly, there is scope to increase public resources for higher education (2 nd challenge) but, this measure alone, is unlikely to cover the additional costs. Given the small share of total costs currently covered by students and their parents – and the high private returns to tertiary education – asking students to contribute more seem sensible (3rd challenge). Yet, two of the challenges (4th and 5th challenge) make this difficult: private costs (including living costs) are already high relative to average incomes in Romania, and, to maintain or increase current enrollment, participation rates will need to increase, but this will involve getting more students from disadvantaged and poor rural areas to enroll. Doing so and getting these students successfully through the system will require substantially more resources and raises the question: who should pay? One answer – supported by the Romanian Constitution, which says “education shall be free, according to the Lawâ€? – is to find the additional resources in the state’s budget. However, given the cost pressures described above, this would require policy makers to divert tax revenue from other public priorities, and/or to raise taxes. Another answer – and in many ways, the one already chosen by Romania – is to ask students and their parents to share part and, preferably a greater part, of the cost of higher education. In many ways, Romanians already seem to have accepted that higher education is an investment with very high private returns, and something worth spending part of their own resources to purchase. As mentioned above, currently, 70 percent of all students in higher education pay a tuition fee. Thus, the discussion on whether or not students should pay for higher education seems no longer to dominate the debate; rather, the question is who should pay, for what, and how much? Romania is not alone in recognizing that public resources, alone, cannot cover all costs associated with higher education. In fact, the rapidly rising cost of higher education is one of the key factors underlying the growth in most countries, including in Europe, of measures to supplement governmental (i.e. tax) revenue with various forms of non-governmental revenue. And a combination of tuition fees, more nearly full-cost fees for formerly governmentally subsidized food and lodging, and the encouragement of fee-dependent private higher education are emerging worldwide as the principal measures of higher education revenue supplementation. Romania is also not alone in recognizing that any shift of costs toward students—particularly if this shift is to be compatible with expanded access to lower income and rural youth—must be accompanied by policies to allow such expenses, whether in the form of tuition or other fees—to be deferred and paid only when the student has graduated and is in the workforce, presumably earning a considerably higher salary in significant part because of the higher education that he or she has received. A policy that allows students to defer paying for their education is the introduction of a student loan scheme. In fact, introducing a student loan scheme seems like an appropriate policy instrument to address a number of the challenges discussed above. First, as discussed, it gives students the means to defer paying for a bigger share of the rising cost of higher education. Second, in principle, a student loan scheme can be a more cost-effective way than grants to support low-income students. In particular, for the same amount of public resources, a student loan scheme can provide financial resources to a multiple of the number of students supported by grants. 28 However, it is important to note that introducing a student loan scheme as an extra policy instrument is unlikely to work on its own in improving participation rates of low income families and other disadvantaged students; existing policy instruments will need tweaking as well and other complementary policies will need to be added. In particular, as mentioned above, it appears that current government financial support for students reinforces the inequitable outcomes observed rather than ameliorating them. And with approximately 400 million RON being used on direct public support for students (excl. the cost of providing tuition waivers – when included, the cost of direct public support is an estimated 2 billion RON), re-thinking how and to whom these funds are allocated is likely to have a substantial impact on equity. In addition, other supportive policies might need to be introduced or expanded to help these groups (e.g. efforts to improve the academic preparedness of low-income students, better counseling in high schools, broader assistance programs to at-risk students when in college, etc). For a policy maker thinking of adding another policy instrument, it is worth noting that international experience suggests that it is more effective to introduce a loan program as part of a comprehensive re-designed student assistance policy, rather than having to go back and fix inefficiencies (and fight the same political battles) later on. This is not to say that the introduction of certain types of loan programs cannot be a strategic first step in a longer-term plan, which it clearly can (and has been in many countries), but rather that it is important to locate the aims and operations of the loan scheme within the broader policy picture from the start. V. Introducing a student loan schemes as an additional policy instrument in Romania A student loan scheme is a policy instrument available to ministries of education in a number of countries around the world. The Romanian education law is mentioning student loans, but so far, no operational scheme was put in place. This section briefly describes: (a) the basics of such an instrument; (b) its possible objectives; (c) the types of repayment arrangements that could be used; and (d) some of the conceptual and political complications in establishing a student loan program. Some of these complexities are treated more fully in a companion piece to this paper: Primer on Student Loans (see footnote 6). The general message of this section is that student loan schemes are far more complex in concept and difficult in their implementation than often assumed. At the same time, a student loan program is the only way through which large numbers of students can invest in their own higher education—which is becoming increasingly important not only in Romania, but in virtually all countries, where the costs of higher education—including not only the high and rising institutional costs of instruction, but also the very substantial expenses of food, lodging, and other costs of students living—are exceeding the capacities both of governments (that is, taxpayers) and parents to fund. 29 V.a Two Basic Principles of Student Lending Two specific implications regarding student loan schemes should be kept in mind regardless of the aim or aims of the loan scheme for Romania. First, student loan schemes are costly. Some of the costs are unavoidable: that is, inherent to student lending because of multiple small loans with long in-school, grace, and repayment periods to borrowers who are exceptionally mobile, who generally lack established credit, and who typically lack collateral and frequently also lack credit-worthy parents to serve as co-signatories. Other costs are essentially discretionary, or the result of political decisions to subsidize the loans with below market rates of interest (in addition to covering the unavoidable costs of small size, high risk, and the like). The so-called unavoidable costs of absorbing or sharing in the inherent risks of student lending may be lessened in Romania by:  keeping the student loan agency as professional and businesslike as possible: that is, free from political patronage or interference;  taking advantage of any foundation or international development agency assistance in the early years to absorb some of the default risk;  requiring parental or other co-signatories from families that are financially able to guarantee their children’s student loan obligations;  requiring low income and other parents who are not sufficiently creditworthy to be a true co- signatory to nevertheless sign an obligation to assist in the skip-tracing of their children who fall behind in their loan repayments. The so-called discretionary costs of future Romanian student loans programs can be lessened by:  rationing, or targeting loans to only those loans that make a difference in participation (in other words, avoiding or minimize lending to students without demonstrated need, which would only allow them to lessen their financial dependence on their parents or increase their standard of student living);  charging interests rates at, or near the government’s borrowing rate;  charging interest on student loans from the origination of the loan, even though interest may be deferred and accrued during the in-school years and an appropriate grace period. In spite of all that might be done to minimize losses and unnecessary expenses, student loans that are generally available — that is, which do not discriminate among potential borrowers by likelihood of default, which would likely discriminate against potential borrowers from lower socio-economic classes or from rural parts of the country, or from certain ethnic or linguistic minority groups — are going to incur substantial annual costs (which are likely to increase each year as participation increases among the riskiest borrowers21). Therefore, a loan scheme should not be implemented under the false assumption that all costs will be absorbed by foundations or donor agencies or that a student loan program, once begun, can ever become self-sustaining by obtaining new loan funds from the repayments on past lending. Second, following from the principle above of student loan programs being inherently costly, at least some of any new Romanian student loan schemes should be linked to revenue enhancement. 21 It is almost a “givenâ€? in countries with low but increasing participation rates that the average student within the participation margin—that is, the students who in the future will be leaving secondary school prepared for and desirous of higher education, but who are currently not so completing—will be, on average, considerably more financially as well as academically needy than the students currently going on to higher education. 30 This principle will need to enter where politically possible and appropriate: for example student loans schemes that support already fee-paying, or private, or advanced professional students. But the essence of student loans is that they allow students to bear some of the high and rising costs of higher education without loss of accessibility and thus enable student contributions to supplement governmental (and parental) revenues for the enhancement of higher educational capacity, quality, and financial assistance. V.b Possible Aims of a Romanian Student Loan Scheme[s] The first decision that is required even before designing a student loan scheme is deciding on the aim or aims of the program. Student loan schemes can have several different aims, not all of them fully compatible. For example, a student loan scheme or schemes in Romania can permit students to enroll and persist who might otherwise be denied these opportunities for want of sufficient funds; to provide the government with a more cost-effective form of student assistance than the current provision of a very limited number of government supported (tuition free) places and limited provision of subsidized lodging and food; to shape behavior in the direction of social needs both in the choice of institutions and programs, as well as in the choice of professions and practice venues; and finally, to enable students to live a little better or more independently from their parents. The workshops and meetings of the World Bank team with Romanian stakeholders in February and April 2008 suggested a number of possible aims, but the following three seem most salient: 1. The first aim would be to put money in the hands of financially needy students in a way that expands participation. Such an aim requires a student loan scheme that is both means-tested (or need-based) and also generally available (as defined above). In other words, loans are made available to all or most students who have a remaining financial need after considering all other sources of revenue, including parents. In this way, the loans would not be available simply to provide a higher student living standard, or to allow students to become financially independent of parents who would otherwise be able to provide at least some financial assistance. As the loan recipients would then be mainly from low income and rural families, whose parents would probably be unable to contribute or even to co-sign for a loan, such a loan program should anticipate a relatively high rate of default.22 In pursuit of increased enrollment and participation, a means-tested student loan scheme or schemes could specifically:  Ease the financial barriers posed by high and rising costs of living for government supported students, especially the relatively low—but desirably increasing—number of students who are from rural and financially needy families.  Ease the even more difficult financial barrier posed by high and rising tuition fees in addition to the rising costs of living that face the fee-paying students.  Make the emerging private sector more accessible to more students with cost-effective assistance in the form of minimally subsidized student loans. 22 Defaults are high in virtually all loan schemes that are generally available: that is, which do not discriminate among potential borrowers by likelihood of default—which discrimination would likely discriminate against potential borrowers from lower socio-economic classes or from single parent families or from certain ethnic or linguistic minority groups. Defaults are also generally high in countries (such as Romania) that are still developing a credit culture in which debts tend to be repaid because a good credit rating is important to virtually all citizens (even to young university graduates). 31 2. A second aim for a student loan program in Romania could be to influence post graduation practice or practice venue. A student loan can be given for the aim of influencing the choice of the student as a graduate to practice a certain profession and/or to practice in a certain target venue: for example, the practice of medicine, nursing, or teaching in a rural district. This is done by granting or even requiring most students to complete with a certain level of indebtedness, portions of which can then be forgiven for each of several years of practice in the target venue. In the case of Romania, where one of the university-related problems is the exodus of higher educated talent to other countries, the debt forgiveness could also be extended simply to all students who returned to Romania with a student debt. Workforce contingent loans assume that professionals will be motivated to do what they would likely not otherwise do (i.e., teach or practice medicine in a remote village for little salary) because of the prospect of debt forgiveness. A second assumption is that student debt forgiveness is more cost- effective (or more politically feasible) than alternatives such as higher salaries, first year bonuses, subsidized housing and transportation, and other incentives that might target public resources to the same end. Research on these questions is inconclusive, and the Romanian government might first experiment with a small, targeted, well-analyzed pilot programs. Nevertheless, such a scheme is, at least in theory, an attractive way to combine the aims of revenue generation, manpower planning, expansion of participation, and rural development. And a pilot program—for example, of student loan repayment forgiveness for practice in some rural provinces (as opposed to the existing scholarship for rural students who agree to return to rural areas to work for a certain number of years) —would be easily implemented and, if designed with minimal subsidization, affordable to the government. 3. A third aim of a Romanian student loan scheme might be to maintain accessibility and participation while implementing a degree of cost-sharing by shifting some of the costs of instruction and/or student maintenance from either the government or the family to the student. Such an aim requires that loan recovery be maximized: that is, that the two principal sources of losses—i.e., of interest subsidies and defaults—be kept to a minimum. The purposes of the enhanced revenue made possible by the cost-sharing, in turn, may be expansion of capacity, enhancement of quality, provision of more targeted (i.e., means tested) financial assistance, a substitution for tax-based governmental revenues, or any or all of the foregoing. In Romania, for example, a student loan scheme could allow an increase in food and lodging fees, thereby allowing a lowering of the governmental subsidization of student maintenance, making up for the increased fees with student loans and shifting the savings to other higher educational needs such as increased financial assistance, increased capacity, or increased quality—but maintaining higher educational accessibility by allowing students, if necessary, to borrow the funds to cover the increased fees. Similarly, a student loan scheme could permit an increase in some or all of the tuition fees paid in the fee-paying tracks—again, by allowing such increases to be borrowed, thus maintaining accessibility. Finally, the universities could be allowed to begin charging fees for some of the most selective, high-cost, and highly remunerative advanced graduate programs, such as graduate management training—again maintaining accessibility, but allowing students to contribute more to the programs with such manifestly high private returns. 32 V.c Forms of Repayment Obligations Student loans may take one of two basic forms of repayment obligation, with variations of each and with a hybrid version also possible. The most important features of alternative student loan schemes are first, the degree of cost recovery (and thus the overall size and the financial sustainability of the program) and second the manageability of the repayment burden (the latter being a complex function of aggregate indebtedness, interest rate, repayment period, and provision, if any, for deferment or forbearance in the event of hardship). The forms of student loans are described more fully in the Annex. In brief, three principal forms are: 1. The Fixed-Schedule, or Conventional Mortgage-Type, Loan: In a fixed schedule, or conventional, mortgage-type loan, the monthly schedule of repayments, the interest rate, and the repayment period are all fixed in the contractual repayment obligation, or loan note. What would vary—mainly according to the income of the borrower (including periods of low or no income, as in unemployment)—would be the annual burden of the payments, which would depend on the initial size of the debt, the interest rate, and the repayment period —and how the resulting monthly payment compared to the borrower’s annual income. 2. The Income Contingent loan: The second common form of student loan is the income contingent (or income-related, or contingent repayment) loan. In an income contingent loan, what would be fixed would be the monthly or annual repayment burden (at least as far as burden is a function of earnings) along with the interest rate (which most borrowers will end up paying in full and which would presumably be the same as in the conventional fixed - schedule option). What would vary—again as a function mainly of the level of income, or earnings—would be the repayment period for those who eventually repay their loans in full or the ultimate cost of the loan for some number of lifetime low earning borrowers. 3. The Hybrid Fixed Schedule-Income Contingent Loan. In a hybrid, or fixed schedule- income contingent loan—the underlying, or default, obligation would be a fixed schedule of payments that would be due unless the monthly or annual repayments exceeded some maximum percentage of monthly or annual earnings. In that event, amounts owed in excess of this threshold would be deferred (and the interest compounded). Borrowers, experiencing a year or two of low income due to unemployment, for example, would pay income contingently during these years, but return to the fixed schedule of repayment obligations when they regained their employment and their earnings. They would have been granted the convenience of automatic deferment of payments—similar to a refinancing—but not a subsidy, as such. However, those borrowers who combined prolonged periods of unemployment or a low paying job with high initial indebtedness might continue to repay their student loans on an income contingent basis, reaching the end of the original underlying repayment period with remaining indebtedness—which at some point would be forgiven as though the entire student loan obligation had been income contingent from the beginning. The form of the repayment obligation is being raised at this point in this discussion paper not because this issue is so critical, but because policy discussions on student loan schemes are so often complicated by confusion between the truly critical elements of a loan scheme —such as the degree of rationing or targeting, or the amount of funds made available, or the degree of discretionary interest subsidization, or the connections to tuition fees and other forms of cost - sharing—and the far less critical, but oftentimes more visible and yet frequently misunderstood nature of the repayment obligation (specifically whether the form is to be fixed schedule, income contingent, or a hybrid combination of both versions). Here are great advantages to a feature of student loan schemes that can provide some form of repayment relief to stud ents 33 who may be having difficulties finding their first job, or have become unemployed, or have chosen a low paying (but socially important) occupation, or who may, for any other legitimate reason, have difficulties in making current payments on their student loans. But such features, as well as any advantage to the lender of having employers collect repayments by deducting them from wage or salary payments can be attached to any form of repayment obligation—as in the hybrid fixed-schedule-income contingent repayment obligation discussed above and, in more details, in the Primer on Student Loans, the companion piece to this discussion paper (again, see footnote 6). V.d Conceptual and Political Complications in Establishing a Student Loan Program. There are certain conceptual and political complications in establishing a student loan program that make difficult the decisions with which a government is confronted. We address briefly six of them: 1. The extent to which the loan is truly a loan versus partly a grant: Since student loans, commonly, are given on more favorable terms than what a commercial bank would be willing to offer, they have an element of “grantsâ€? embedded into them. It is possible to distinctio n between a true loan—that is, the portion of the sum originally borrowed that will be fully repaid, which can also be viewed as the discounted present value of the reasonably anticipated stream of repayments—and an effective grant, or the present value of the stream of subsidies. The effective grant is thus the difference between the amount borrowed and the true loan. The mix of true loan and effective grant is a function of the spread between the interest rate to be paid by the student borrower and the cost of money to the lender, which, at a minimum, would be approximated by the government’s cost of borrowing. 2. Loans as assets or expenditures: That portion of a student loan that is a true loan is an asset to the lender. To the degree to which the amounts to be lent to students come from government budgets and may then appear as expenditures, the impact on the true operating budget may be overstated by the amount that is a true loan, which can be more appropriately viewed as a governmental investment and carried on the government’s books as assets than as expenditures. In fact, if the asset value is considerable, the government (or the student loan agency of the government) can either sell the loans and recover the present value of the repayment stream immediately, or can securitize the assets and sell its own paper in the capital market, backed by the student loan assets. Thus, there is great benefit to maximizing the asset value of student loans, which enables a given volume of taxpayer funds to put far more money in the hands of students than the same taxpayer funds given as grants, stipends, or tuition fee discounts. If the government (or a governmental or public agency) does not actually make the student loan but only guarantees loans made by banks or other financial institutions, the accounting is essentially the opposite of the example above; that is, there may seem to be no expenditures as such (assuming no subsidization of interest other than the implicit subsidization of the guarantee), but there would be major contingent liabilities on a proper governmental balance sheet because of the risk of default—or alternatively, a major operating expenditure sufficient to fund a reserve large enough to cover any future losses. 3. The costs of student loan schemes to governments: Student loans are costly to governments. The costs are in two forms: (1) interest rate subsidies and (2) guarantees against default. The cost that is borne by government in the form of a guarantee against default loss, at least on student loans that are generally available (i.e., not rationed for low risk student 34 borrowers only), is an unavoidable governmental obligation, although the guarantee cost can be substantially lessened by sharing the cost of losses with parental co -signatories, foundations, or even with higher educational institutions. On the other hand, the cost of governmental subsidies—i.e., the spread between the interest rate to be paid by the student borrower and the cost of money to the governmental lender—is essentially discretionary and is a political decision made at the loan scheme design phase of what interest rate to place on the student loans. And the reason that student loans programs are as costly as they are in so many countries is the tendency of politicians to make financial obligations oftentimes with little consideration of the future implications: in the case of student loan schemes, a stream of annual interest subsidies that will be paid for by future governments, future politicians, and, more to the point, by future taxpayers. 4. The politics of student loan schemes Student loan schemes are oftentimes politically charged by their association with cost-sharing and with a shift of some of the costs of higher education that may currently be borne by governments to parents and/or to students. As mentioned above, one of the aims can be to preserve accessibility in the face of policies that might shift some costs to parents and/or to students. On the other hand, student loans are needed whatever the degree of cost-sharing, as taxpayer funds, at least in market economies, are never sufficient to cover all of the costs of higher education (without severe and inequitable rationing or losses in quality). But opposition to student loans by student unions, where such opposition may be found, is rarely, if ever, directed to the provision of loans themselves, but rather to a presumption that student loans may be linked to a lessening of the taxpayer subsidies that the students are now enjoying, or are attempting to secure. 5. Irrational borrower behavior: Student loan schemes are complicated by borrower behavior that may frequently depart from the economist’s notions of perfect consumer, or borrower, rationality. Furthermore, socio-economic class or ethnicity or region may affect these departures from economic rationality. For example, some cultures discourage borrowing. Some cultures/religions—Islam in particular—discourage the charging, or payment of conventional interest (although allow other ways to reward those who make their savings available to investors). At the other extreme, students from cultures where credit is virtually ubiquitous (the United States is a good example) may have the opposite irrationality: that is, being seemingly oblivious to the implications of mounting debt and underestimating the burden of excessive borrowing—or borrowing so-called private loans at high rates of interest when far more affordable rates are available on governmentally-sponsored student loan schemes. 6. Financial sustainability and risk: While most experts agree on the potential of student loans, they also acknowledge that student loan programs in many countries have not been financially sustainable—at least not at levels high enough to promote widespread participation. For financial sustainability, student loan programs need not only to limit the subsidy costs of student lending to levels that the government can afford (as mentioned above in point 3), but they also need to address the capital constraints that occur when only government resources are being used to make the loans. This second dimension of financial sustainability—the ability to tap private capital markets rather than government budgets for the student loans—is complex. In theory, as well as practice, the amounts originally lent can come largely or entirely from banks and other private sources of savings, yet obligate the government to very considerable subsidies, both in the form of the in-school interest subsidies and subsidies during repayment, and also obligate the government, as guarantor, to pay the private lender for loans in default. Therefore, what may seem like private lending in the first instance, with no immediate impact on the governmental 35 budget, can in fact be indistinguishable in proper accounting terms from a considerable public expenditure, albeit in the form of future and contingent liabilities, rather than outright current expenditures. On the other hand, a loan requiring little, or no subsidization (that is, with a flow of repayments sufficient to cover the cost of money plus the cost of administration and servicing) and little risk of default to the lender (because of sufficient collateral by either the borrower or a co-signatory) could be made directly by the government, with taxpayer money rather than by a bank or other private lender and be treated not like an expenditure at all, but rather like a government investment. In theory, then, the financial sustainability of a student loan program could by assured by eliminating, or at least minimizing governmental subsidization and somehow minimizing the element of risk. Tapping into the private capital market, would then be easy, whether the student loans were originated by banks, or were originated by a government agency, or even by a university and then sold to banks or other institutions that comprise the private capital market. The fact remains that there is always an element of risk, at least in generally available student lending (that is lending without regards to parental wealth, academic program, or other attributes by which a potential student borrower might be risk rated). This risk is beyond that which can reasonably be expected to be covered by lenders or parental co-signatories, and which then presents governments with some element of at least contingent liabilities—the present values of which are akin to ordinary governmental expenditures and which, therefore, become impediments to financial sustainability. The major factors associated with the risk of student lending include: ï‚· The absence of collateral: The fundamental riskiness of student loans is due to the fact that unlike home mortgages, auto loans, and some other forms of commercial debt, a student loan provides nothing for the lender to repossess in the event of non- repayment. ï‚· The absence of a general credit culture. In the highly industrialized countries, credit histories, maintained by credit agencies, play a vital role in making cost-effective credit available to good risks—and conversely, in keeping bad credit risks from borrowing and raising interest rates to everyone. Students typically have no credit histories, but when they enter repayment as young adults, they quickly learn the importance of maintaining a good credit history, without which automobile and home purchases and even credit cards may be out of reach. Student borrowers in countries lacking such a credit culture may be more prone to default, as they may perceive less of a need for a good credit history. ï‚· Students misunderstanding the nature of the repayment obligation: This may be due in part to the immaturity of the student at the time of borrowing and/or to the considerable length of in-school time between the original borrowing and the supposed beginning of the repayment obligation. Misunderstanding is also more likely when the original loan never passes through the student’s hands, as it does in the case of an automobile loan or in most commercial credit, but rather is simply recorded as a future obligation upon a student’s matriculation. ï‚· The extreme mobility of students for a period of time after leaving the university: Students typically move around in search of first employment, or simply the lure of young wanderlust. For several years, they may not have what they consider regular employment and typically do not own a home or real property. They may be traveling abroad, leaving no forwarding addresses. ï‚· Other country-specific factors: Other factors could include such factors as extreme political instability. 36 Given the risky nature of student loan enterprises, it is not surprising that governments have to play a role. A student loan program can be 100 percent public, with the government playing all roles from origination to collection, such as in most developing countries. However, the two functions that most require the participation of government are guaranteeing repayment (at least if the student loans are to be made generally available) and subsidizing interest (if the interest rate to the student borrower is to be less than the rate for general, medium -term, unsecured, consumer debt). Evidence from around the world has shown that the most successful loan programs have the government retain these two core processes and outsource some or all of the others (United States, UK, Japan). Given the risks inherent in generally available student loans, it is unlikely that governments can ever get out of the student loan business altogether. Nevertheless it is important for governments to develop a long term strategy for their loan program; one that may in the short run rely exclusively on government funds, but in the longer term, looks for ways of working with the private sector to generate additional capital and risk sharing. In this context, mention must be made of the serious efforts being undertaken by the Roma nian American Foundation to create a pilot student loan scheme in Romania that would guarantee loans provided by the Raiffeisen Bank to Master’s students. The Foundation is still conducting a qualitative and quantitative market study on which they will base their next steps, but have agreed to share all of their findings with the MERY. Their groundbreaking work to involve a private bank in student loans (albeit particular types of loans with far less risk of non - repayment than generally available loans) is extremely helpful to the ministry’s own efforts. The track record that they manage to accumulate with their program can be useful in convincing other private sector banks to become involved as well. 37 VI. Estimating the fiscal cost of introducing a student loan scheme The fiscal cost to the government of student loans is conceptually more challenging to estimate and discuss than, say, the fiscal cost associated with a student grant scheme. The main difficulty is that loan disbursements – whether done directly by the government or via private banks – are not proper expenditures such as grants. Rather, loans are financial claims, which the government (or a private bank) has on students. Alongside the disbursement, there is an expectation of repayment at an agreed upon interest rate over an agreed upon period of time. Thus, it would be inappropriate and misleading simply to focus on the size of loan disbursements which, depending on the size of the loan scheme, could be very large. For instance, offering a loan program of, say, 500 euros per month to, say, 5 percent of the currently enrolled students (i.e., 50,000 students) would imply annual disbursements of 300 million euros, or almost a third of the public spending on higher education in 2008. If properly designed and managed, a large proportion of these 300 million would, one day, be repaid. Conceptually, the correct fiscal cost to focus on is a somewhat abstract notion, namely, how much would it cost the state to contract someone (most likely a private bank) to disburse and collect (down the road) an agree-upon amount of money under pre-specified interest rates and repayment schedules? As will be discussed below, it turns out that it is possible – although only very crudely – to estimate what a private bank would need to be paid to carry out such a task. In the discussion below, we will refer to this cost as the “true costâ€? of a loan scheme. Obviously, a private bank would not need any fiscal resources if they could charge an interest rate which could fully cover the relatively higher risk associated with lending to a student and all their costs of running such a scheme (e.g. the state does not need to subsidize a bank to stimulate mortgage or car loans). It turns out that the true cost of a loan scheme is a function of a number of parameters, each of which can be examined separately and assessed in terms of their impact on costs. The parameters which determine the true cost are: (i) the interest rates (relative to market interest rates) charged during the years in school, possible grace period, and repayment period; (ii) the length of the repayment period; (iii) an estimate of the rate of default (such estimate would take into consideration all, if any, clauses allowing for debt forgiveness); (iv) the cost of administering the loan;23 In general, a loan with a very subsidized rate of interest (i.e., an interest rate that is below the prevailing market interest rates) and a very long repayment period will be far more costly to the government than an unsubsidized loan with a short repayment period. Although at the risk of getting into too technical details, it is worth noting yet another conceptual difficulty of calculating the fiscal cost of a student loan scheme. Similar to the exercise of evaluating other investments with cash flows in the future, the only way to assess the true cost of a student loan scheme is to convert both disbursements and expected repayments into “present value 23 Whether the administration is done within a state agency, or by a private bank, there will be administrative costs of running the program (e.g. staff costs, computer costs, etc). For the purpose of estimating the true cost of a student loan program (i.e. getting an estimate of what a bank would charge to take on the responsibility of running the program), it is important to estimate all costs associated with the program (even if civil servants are paid less than employees at a bank, and/or are already working in a ministry, and/or could be diverted from doing other activities). 38 termsâ€?.24 When doing so, one is recognizing that, say, 5,000 RON disbursed today will not have the same purchasing power and value to future citizens of Romania 10 or even 15 years from today. In the case of a student loan scheme, the present value of disbursements will be known with certainty, but the present value of repayments can only be estimated, based on the set of parameters given above. In fact, the present value of future repayments can be calculated using different sets of assumptions and reflecting different policy decisions to get a sense of what the government will actually spend on a certain type of loan. To provide policy makers with a likely range of the fiscal costs involved in introducing a student loan scheme in Romania, several different student loan scenarios were examined. The scenarios below do this by first finding what is owed at the time repayment begins (which is the amount lent, if no interest is accrued during the in school and grace periods, but will be higher than the amount lent, if interest has been compounded) and then finding the future value of this amount under different sets of assumptions regarding the interest rate level. The future value is then discounted back to the present day using the country’s discount rate to get the present value of the future payments, i.e., what the government can expect to get back in today’s terms. The scenarios include a number of design decisions: ï‚· What should be the length of the grace period during which no repayments are due? ï‚· What should be the interest rate for the in school and grace periods? ï‚· What should be the interest rate for the repayment period? ï‚· What should be the length of the repayment period? ï‚· What should be the amount of the loan? They also include two parameters that are essentially the result of loan administration: ï‚· What is the cost of administering the loan? ï‚· What is the default rate? Once values are plugged in for the design parameters and values are assigned for the implementation parameters, the present value of future loan repayments (less the cost of administering the loans and defaults) can be calculated to find the degree to which the repayments offset the cost of the loans. 24 Present value calculations are widely used in the finance literature to enable financial planners to compare cash flows occurring at different times in the future. By converting cash flows occurring in the future into “present valueâ€? terms, they are converting the future cash flows into the same unit of measurement (i.e., in today’s value of RON). 39 Table 9: Estimate of the fiscal cost of different student loan scenarios Scenario One Scenario Two Scenario Three Heavily Subsidized Moderately Subsidized Minimally Subsidized Loan Amount ï‚· State 6,458 RON 6,458 RON 6,458 RON Sponsored Student ï‚· Fee paying 11,070 RON 11,070 RON 11,070 RON Student ï‚· Private Student 11,070 RON 11,070 RON 11,070 RON Length of grace period 2 years 2 years 2 years during which no repayments owed Interest rate during in 0 6.57% 9.4% school year and grace period Interest rate during 9.4% 9.4% 9.4% repayment Length of repayment 15 15 15 period Number of students ï‚· State 5,011 5,011 5,011 Sponsored Student ï‚· Fee paying 5,011 5,011 5,011 Student ï‚· Private Student 8,640 8,640 8,640 Present value of loan 183.5 million RON 183.5 million RON 183.5 million RON disbursements Present value of student 50 million RON 68 million RON 78 million RON repayments Recovery rate 27 percent 37 percent 43 percent Table 9 shows the different amounts that are recovered by student loan repayments depending upon the loan parameter decisions made by the government. In all of the scenarios, it is assumed that the loans carry an administrative cost of 2 percent, a default rate of 30 percent and that the country’s discount rate is 12 percent. As a proxy for the number of “needyâ€? students (i.e., the recipients of this type of loan), all three scenarios use the current number of students receiving a “social scholarshipâ€? (discussed above in Box 3), approximately 5,000 students. All of the loans cover tuition fees and living expenses for fee paying students in the public sector and private sector students (11,070 RON) and living expenses only for state supported students (6,458 RON). They all have an average in school period of three years and a grace period (during which no payments are made by the students) of two years. The first scenario models the introduction of substantially subsidized government loans (no interest accrued during in school years, which given inflation, implies a negative real interest rate and two year grace period; government’s borrowing rate during repayment) for needy students in the public and private higher education sector (undergraduate and graduate) to address upfront financial barriers to participation. The government would use its own resources and/or borrow capital at low 40 interest rates from international organizations. The loans would be means-tested using existing or new means testing mechanisms. Risk of defaults would be covered by credit-worthy cosignatories, when available and the government when cosignatories are not available. With such large grant elements built into the loan (and large provisions for defaults), only a small fraction of the loan amounts disbursed would be recovered. In fact, only an estimated 27 percent of the loan amount would be recovered. The second scenario includes the introduction of moderately subsidized loans (zero real rate of interest during in school years i.e., given inflation, students will be paying a positive nominal interest rate and grace period; government borrowing rate during repayment period). In this case, the recovery rate would be higher at 37 percent given that interest was accrued during the in school years and grace period. Finally, the third scenario includes the introduction of minimally subsidized, need-based loans. Interest is set at government borrowing rate for the life of the loan and accrued starting at disbursement. Given the minimal subsidization the repayment rate is a higher 43 percent. Offsetting Costs Student loan programs can be introduced with or without simultaneous changes in other parts of the higher education finance and student financial assistance system. Capital for the loan program can come from additional government money, government borrowing or donors, from the reallocation of existing money in the higher education budget, or from the introduction of cost sharing measures. In Romania, examples of the latter two include: reductions in government subsidies for transportation and/or student food and lodging costs (at present 28.7 million RON and 168 million RON respectively); increased fees to full-cost recovery for fee-paying students (at present, the amount of instructional costs covered by fees varies dramatically from institution to institution), increases in tuition fees for advanced professional programs (at present, state supported students pay no tuition fees and the tuition fees charged to fee paying students are comparable to those charged in undergraduate programs) and the introduction of moderate tuition fees for all. While politically more difficult in the short term, international experience suggests that it is more effective to introduce a loan program as part of a comprehensive re-designed student assistance policy, rather than having to go back and fix inefficiencies (and fight the same political battles) later on. This is not to say that the introduction of certain types of loan programs cannot be a strategic first step in a longer-term plan, which it clearly can (and has been in many countries), but rather that it is important to locate the aims and operations of the loan scheme within the broader policy picture from the start. 41 VII. Conclusions and Recommendations Student loans are a policy tool that properly designed and implemented could contribute to Romania’s higher education objectives in terms of enrollment expansion, income generation and program/venue choices. Nevertheless, even the best implemented programs can be costly if excessive subsidies are built into their design. In the view of the World Bank team, a student loan scenario that simply is implemented on top of the existing higher education financing structure (scenario 1) with no offsetting additions to financial assistance revenue would a) not be financially viable in the long run and b) would have unacceptable opportunity costs in view of the substantial government investment that is needed in capacity and quality. Moreover, focusing on student loans without addressing other issues of educational access and finance will not be effective in increasing participation or reducing inequities; existing policy instruments in higher education also need tweaking to address these challenges. In particular, existing government financial support for students is likely to need adjustments since it appears to reinforce the inequitable outcomes observed rather than ameliorating them. Moreover, equity at higher education is unlikely to improve unless more focus is directed on the quality of the high school education which disadvantaged students receive. The paper, therefore, has two main recommendations: first, introducing a minimally subsidized loan scheme (see below) is advisable. Second, redesigning existing student financial support scheme would be critical to help improve the inequitable outcomes observed in higher education. In terms of the first recommendation, a minimally subsidized loan scheme that is integrated into the Romania’s student financial assistance policy could: ï‚· increase higher education accessibility by putting additional funds in the hands of needy students; ï‚· allow some portion of the costs of instruction and student maintenance, or costs of living to be shifted to the students to be paid as he/she enters the workforce. In this way student loans would provide additional revenue to higher education for the purposes of enhancing capacity, quality and participation; ï‚· influence post graduation behavior in ways that the government desires using loan forgiveness. In order to do this, however, the loan scheme would have to: ï‚· Be part of a comprehensive package of cost sharing with loan amounts sufficient to cover tuition fees and living costs (less grants and expected parental contributions); ï‚· Have reasonable repayment expectations with repayment obligations spread over enough time to be manageable; ï‚· Have automatic deferral options that are activated in the event of unemployment, prolonged illness or loss of work, maternity, etc. ï‚· Be designed to minimize costs by keeping the base interest rate close to the government’s own borrowing cost, minimizing servicing and collection costs and minimizing defaults by assuring that the students view the loans as real obligations, by requiring some form of repayment plan to be agreed upon before graduation, by engaging in good collection practices and possibly by requiring or encouraging repayments to be made through employer deduction at the time of wage and salary payments; ï‚· Be designed to share risk between government and cosignatories. Co-signatory requirements may need to be limited to moral persuasion and loss of credit reputation so as to avoid losing those parents who have insufficient assets to pledge. 42 In summary, effective student loan programs are possible and can enhance both institutional viability and student accessibility and participation. However, student loans are complex, and require both proper design and good execution. Unfortunately, too many student loans programs have failed, both for poor design and poor execution. It is hoped that this paper can be a small contribution to the policy planning required to begin a new student loan program in Romania. In terms of the second recommendation, to improve equity in higher education, government financial support to students could be redesigned in such a way that the bulk of fiscal resources would be spent on the financially weakest students, and only limited resources spent on the resourceful students (e.g. as an incentive for the very best-performing students). Doing so, would enable fiscal resources to matter more by encouraging desired outcomes which, without the fiscal support, would otherwise not have happened. Under the current financial support schemes, the bulk of fiscal resources end up supporting students who would most likely have pursued a tertiary degree even without the public support. Moreover, fiscal resources end up supporting students who are more likely to come from resource-strong families who have motivated them, paid for tutoring, and provided them with an environment conducive for learning. Next steps: should policy makers decide to introduce a student loan scheme, the following next steps will be needed: 1. Issue a government decision to disburse first student loans by, say, October 2009. The decision should contain a broad estimate of the additional fiscal resources to be allocated for this purpose in the first years of the program. 2. Assign staff to work full-time on designing student loan scheme, addressing each of the issues raised in sections V and VI of this paper (e.g. what is the loan scheme(s)’s primary objective, and what monitoring arrangements will be introduced to assess whether those objectives are reached? Who will be eligible to receive the loans? What will be the terms of the loans? What are the repayment arrangements? etc) 3. Shift the focus beyond policy design questions to implementation arrangements, including what will be the role of the main agents in any student loan scheme (governments and ministries; public agencies; banks and other capital sources; universities and colleges, parents and other co-signatories; and collection and servicing agents) and how will the main functions (such as setting eligibility terms and loan term parameters, originating loans, bearing risk of default, subsidizing loan, providing capital, and servicing and collecting loans) be carried out? 43 References ï‚· Coomans, Gery. (2005). The Demography/Education Squeeze in a Knowledge-Based Economy (2000-2020). Technical Report EUR21573. Spain: European Communities Joint Research Centre. ï‚· EurActiv. (2008, April 25). Romanian skilled labour shortage ‘highest worldwide’. ï‚· Eurybase – Romania (2006/07). ï‚· Higher Education Information System (HIS). (2008). Social and Economic Conditions of Student Life in Europe. National Profile of Romania. Eurostudent III 2005-2008. ï‚· Hua, Shen and Adrian Ziderman. (2007). Student Loans Repayment and Recovery: International Comparisons. Ramat Gam, Israel: Bar-Ilan University. ï‚· Johnstone, Bruce and Marcucci, Pamela: Primer on Student Loans, mimeo, State University of New York at Buffalo. ï‚· Korka, Mihai (2008). Implementing the Modern Approach of Quality Assurance in Romanian Higher Education ï‚· Ministry of Education and Research, National Council for Higher Education Financing. (2006). Methodology of State-Budgeted Base Finance Distribution by Higher Education Institution in 2006. Synthesis. ï‚· Popescu, Cristian C., Laura Diaconu, and Andrei Maxim. (2007). Human Capital and Migration Costs for Romanian Economic Development. ï‚· Romania Economy Watch. (2008). Romania at A Glance – January 2008. ï‚· Ziderman, Adrian. (2004). Policy Options for Student Loan Schemes: Lessons from Five Asian Case Studies. Bangkok: UNESCO and Paris: International Institute for Educational Planning. 44 Annex 1: Enrollment rates in higher education Annex table 1: Enrollment, population and enrollment rates in higher education Total enrollment in higher education (ISCED 5 and 6), in public and private, part-time and full-time, by age 1998 1999 2000 2001 2002 2003 2004 2005 2006 18-24 294,112 336,314 376,412 435,417 466,681 482,094 500,871 524,832 573,647 25-29 51,095 54,214 58,014 69,806 81,734 110,061 118,362 132,629 160,925 30-34 7,760 9,013 9,795 15,373 18,228 27,169 29,244 31,055 37,834 35-39 7,623 8,179 8,400 12,556 15,578 24,587 37,241 50,290 62,563 40+ 0 0 0 0 0 0 0 0 0 Total 360,590 407,720 452,621 533,152 582,221 643,911 685,718 738,806 834,969 Total population, by age 1998 1999 2000 2001 2002 2003 2004 2005 2006 18-24 2,714,264 2,747,370 2,699,659 2,622,943 2,388,282 2,351,425 2,324,741 2,322,886 2,315,132 25-29 1,905,101 1,796,998 1,802,186 1,825,436 1,718,873 1,749,525 1,789,774 1,783,457 1,784,996 30-34 1,423,894 1,622,293 1,769,083 1,897,166 1,939,939 1,834,259 1,714,709 1,688,340 1,681,325 35-39 1,438,968 1,368,607 1,312,834 1,258,476 1,195,219 1,376,174 1,562,251 1,698,775 1,818,272 Total 22,526,093 22,488,595 22,455,485 22,430,457 21,833,483 21,772,774 21,711,252 21,658,528 21,610,213 Enrollment rates (%) 1998 1999 2000 2001 2002 2003 2004 2005 2006 18-24 10.8 12.2 13.9 16.6 19.5 20.5 21.5 22.6 24.8 25-29 2.7 3.0 3.2 3.8 4.8 6.3 6.6 7.4 9.0 30-34 0.5 0.6 0.6 0.8 0.9 1.5 1.7 1.8 2.3 35-39 0.5 0.6 0.6 1.0 1.3 1.8 2.4 3.0 3.4 Source: Eurostat 45 Annex 2: List of experts met during the paper preparation period In addition to the workshops, the team met with a variety of experts and stakeholders to gather information and get their input on a possible student loan program in Romania including: ï‚· Professor Remus Pricopie, Secretary of State, MERY ï‚· Professor Gabriella Pasztor, Secretary of State, MERY ï‚· Professor Adrian Curaj, Counselor of State, Romanian Government ï‚· Professor Mircea Miclea, Head of the Presidential Commission for Education and Research ï‚· Professor Ecaterina Andronescu, Rector of the University Politehnica of Bucharest ï‚· Professor Radu Damian, President, CNFIS ï‚· Gabriela Jitaru, Head of Higher Education Funding Department, CNFIS ï‚· Ana Cristina Moise, Expert, CNFIS ï‚· Costoiu Mihnea, General Manager of the University Politehnica of Bucharest ï‚· Doina Crisan Habean, Director General of the Bucharest Academy of Economic Studies ï‚· Professor Lazar Vlasceanu, Deputy Director, European Centre for Higher Education, UNESCO ï‚· Professor Adrian Miroiu, National School of Political Studies and Public Administration, Faculty of Political Sciences, Bucharest ï‚· Oana Almasan, National School of Political Studies and Public Administration, Faculty of Communication and Public Relations, Cluj ï‚· Professor Dumitru Miron, Prorector, Bucharest Academy of Economic Studies ï‚· Professor Ioan Alecu, President, University of Agronomics and Veterinary Medicine, Bucharest ï‚· Student representatives of the National Union of Students of Romania and the National Alliance of Student Organizations in Romania. ï‚· Professor Ovidiu Folcut, Prorector, Romanian-American University, Bucharest ï‚· Anca Harasim, Executive Director, The American Chamber of Commerce in Romania ï‚· Doina Ciomag, Executive Director, Foreign Investors Council ï‚· Roxana Vitan, Vice President, Romanian-American Enterprise Fund, Romanian-American Foundation ï‚· Mihaela Ciobotea, Consultant, Romanian-American Enterprise Fund, Romanian-American Foundation 46 Annex 3: Examples of Current Student Loan Programs Drawing on, and providing examples of, the aforementioned principles, some current student loan programs are briefly described below. Additional examples can be found at a website maintained by the authors’ of this discussion paper.25 Australia: Australia introduced its Higher Education Contribution Scheme (HECS) in 1989 as a tuition fee that could be deferred for all Commonwealth-supported students and repaid as an income contingent loan. In 2005, HECS was folded into the Australian Higher Education Loan Programme (HELP). Commonwealth-supported students are entitled to the HECS-HELP loan scheme, which covers the full amount of the tuition fee within three bands as established by the university up to limits set by the government. The upper limits in 2006-7 have been set at A$4996 [$3540 using 2006 ppp estimate] for Band #1 (humanities, social and behavioral sciences, languages and visual and performing arts); A$7118 [$5,050] for Band #2 (engineering, science, computer science, and business/economics); and $A8333 [$5,910] for Band #3 (medicine and law). Up to 20 percent of the tuition due is discounted for paying “up front.â€? The interest rate, as in Sweden and the UK, mirrors the rate of inflation—that is a zero real rate of interest. Repayments are income contingent on annual incomes above A$ 38,148 [$27,055]. Rates range from 3 percent to a maximum of 8 percent on annual incomes in excess of A$ 64,999 [$46,100]. Repayments due are collected as income surtax by the employer or are paid along with estimated or year-end taxes due. There is no forgiveness after a certain age or passage of years since the borrowing took place. According to the definitions above, HECS is not a true graduate tax as individual accounts and balances owed are maintained on each borrower. However, enlistment of the national tax system gives HECS the appearance of a graduate tax and assures both a low administrative cost of servicing as well as a very low default rate. (There is also a loan scheme, labeled FEE HELP and described in Chapman and Ryan [2002], for non-Commonwealth supported—that is, non-HECS eligible—students collected in the same way and with zero real interest, but without the in-school and grace period interest subsidies. China: China’s loan programs have undergone many modifications since their experimental beginnings in 6 cities in 1999. The Government Subsidized Student Loan Scheme (GSSLS) as modified in 2004 provides student loans in amounts up to Y6000 [$109] a year to needy students (officially acknowledged to be 20 percent of the enrollment). Interest rates are paid by the government during the in-school years. Borrowers pay one-half of the commercial interest rate after graduation, which is deferred (but not forgiven) for up to two year’s grace period. Repayment periods are 6 years, which is an increase over the prior 4-year repayment period that required far too high monthly payments. Participating banks disburse the loans, and the university, the government, and the bank share the risk. Co-signatories are not required for the GSSLS. There is also a non-subsidized student loan program, the General-Commercial Student Loan Scheme (GCSLS), available for children of the more affluent families, requiring a parental co-signatory (Shen and Li, 2003). Japan: The newly created independent administrative institution, Japan Student Services Organization (JASSO), administers the recently revised student loan system. The system is made up of two types of student loans: the first class scholarship loan that is interest free and awarded based on merit and need, and the second class scholarship that is interest free during in-school 25 International Comparative Higher Education Finance and Accessibility Website,