Student loan programs are among the most complex, controversial, frequently misunderstood, and yet potentially important elements in the financing of higher education. Their importance stems from the increasing prominence of cost sharing—meaning the shift of at least some higher education costs from governments and taxpayers to parents and students—on the public policy agendas for higher education in most countries. And yet the international higher education policy landscape is littered with loan programs that have either failed outright, or failed to accommodate the devilishly difficult policy balance between expanding higher educational participation and accessibility, while simultaneously expanding real cost recovery from students.

This paper will look at student loans for higher education from an international comparative perspective, looking especially for the essential elements of lending and borrowing for higher education that lie beneath some of the more visible features of certain student loan programs, both in theory and in practice. We will examine especially the challenges of student lending in low-income, or "less industrialized," countries, as well as countries "in transition" from predominantly state-owned means of production and governmentally-controlled economies to market-oriented economies with substantial private ownership. We will search for possible explanations for the difficulties that many student loan programs in such countries have had, and suggest some principles for better accommodating the twin goals mentioned above. (In addressing much of this analysis to so-called developing and transitional countries, we do not equate the economies cultures, or especially the higher education systems of a typical "less industrialized" or "developing" country with those of the former Socialist countries of East and Central Europe or with Russia or most of the other states emerging from the former Soviet Union. However, most of these countries have in com on such features--relevant to student loan programs--as academically selective public sectors, little history of "cost sharing," severe limitations on public taxing capacity, limited and only recently-developed programs of need-based financial assistance and/or student lending, relatively undeveloped private capital markets, and uneven use of private credit.)

I
Student Loans as Public Policy: Governmental Participation

The emphasis will be on student loans that depend in some way on government or public policy—mainly to cover, or subsidize, some portion of the costs of lending, but also to establish a legal and administrative framework for the origination and collection of student loans, or otherwise to intervene to bring (student) borrowers together with lenders, or savers. As such, this inquiry is only slightly interested in the borrowing and lending for higher education that arises out of strictly private transactions between lenders (savers) and student borrowers. Such lending and borrowing in support of higher education goes on all over the world. But it goes on only at a very small scale because the buying and selling of claims on goods and services—that is, the borrowing and lending of money--depends so fundamentally on the certainty of repayment. Without this certainty, the lender will either not part with his claims at all, or will do so only at a very high price (i.e. very high interest rate) so that the premiums from those who do repay can cover the inevitable losses from those who do not. The certainty of repayment, in turn, depends either on the reputation, or credit-worthiness, of the borrower, or on the pledge of recoverable assets, or collateral, equivalent to the value of the loan, that the lender can claim (with the support of the law) in the event that the borrower does not repay.

A fundamental problem with lending to students—and especially to needy students in low-income countries— is that they have neither established "credit-worthiness" nor collateral. An exception may be students of, say, medicine, who are almost certain to become at least relatively affluent and who are also likely to be in need of a good credit rating, who therefore may be considered acceptable risks even with no collateral. Otherwise, lending to students in the absence of governmental or other external assumption of risk almost always requires the co-signature of a parent or friend who has pledgable assets, or collateral, to assure repayment if the student borrower either cannot or will not repay. In such a case, the lender (whether a private bank or a governmental loan agency) may be viewed as lending not to the student, but to the parent or other co-signatory from whom the lender will collect in the event of default by the student borrower. And according to the perspective that the essence of lending is bearing the risk of non-repayment on the part of the borrower, the real lender in such a situation may be viewed as the parent or other co-signatory who is putting his or her assets (not only reputation, and creditworthiness, but oftentimes a house or farm or business or stocks or savings accounts) at risk.

In short, although such strictly private borrowing and lending—with neither public assumption of risk nor public subsidy of the interest rate--may have higher education as its ultimate purpose, it will not be a large or important player in the expansion of higher educational opportunities to students who would not be able to attend (or not be able to attend a particular institution or to attend for a particular program of study) in the absence of the loan. The student loan programs that have become integral parts of higher education finance in, for example, the US, Canada, Sweden, the other Scandinavian nations, Australia, Germany, the Netherlands, and very recently the UK, all entail general availability, in turn requiring a substantial element of government participation.

General Availability. The general availability of student loans means loans that are avail-able to all eligible students. This does not mean that there cannot be restrictions and rationing consistent with the broad purposes of higher education—e.g. by "need," or academic performance, or program of study. But this rationing should not be solely for the purpose of limiting the financial exposure of the government or lender: for example, restricting loans only to those students able to produce parental co-signatories willing and able to pledge collateral, or to students who possess other attributes (such as academic merit or being a student of medicine) that are mainly proxies for "credit-worthiness." Thus, if the principal lender—whether private or governmental--chooses to limit its financial exposure by requiring parental co-signatories wherever possible, the criterion of general availability requires there to be alternate, govern-mentally-sponsored, ways for otherwise eligible students from low–income, non-creditworthy families to receive credit.

Governmental Participation. There are four principal ways for the government to participate in generally available student lending. These are:

1. Bearing all or at least a significant part of risk. Bearing risk is fundamental to general availability, as asserted above. This may be done through government guarantees to private lenders, or through the government actually being the lender (in which case public funds will be put directly at risk).

2. Subsidizing the rate of interest, or the cost of lending paid by the student borrower. Even without considering cost of default (which, of course, needs to be covered from somewhere), the rate of interest needed to purchase money from savers, plus the cost of administration and collection, may be higher than public policy would wish to charge the student borrower. If the government is the lender, subsidization could bring the rate of interest charged to the student borrower down below the rate at which the government itself must pay for funds in the capital market. If the government is merely the guarantor of private lending, subsidization is more likely to mean paying a portion of the interest costs to the private lenders, at least while the borrower remains a student (to preclude the need for accruing interest on the original debt).

3. Absorbing (and effectively "hiding") some of the costs of administering the loan program. There are administrative costs entailed in originating the loans, determining student’s eligibility, collecting the repayments (loan servicing), and the particularly costly task of collecting loans in arrears. These are all costs that, in private lending, would need to be recovered from the borrowers and thus reflected in the interest rate. The govern-ment, however, can absorb some of these costs within the publicly-supported university administrations (such as offices of student services or of financial assistance, or the university business office itself) as well as within the larger governmental costs of collecting personal income taxes and pension contributions, thus minimizing, and to a degree "hiding" but not altogether avoiding, such costs.

4. Employing for student loan collection the potentially powerful machinery of the governmental tax and/or pension collection systems. This may be seen as an extension of the 3rd role, above: government can potentially lower defaults, along with the direct costs of servicing and collection, by collecting loan payments along with taxes and/or pension contributions. (A variation on this fourth role—and to a similar purpose--is using the force of law to require employers to withhold and remit student loan repayments at the point of wage or salary payments, like tax or pension withholding, but technically and legally separate from state tax collection.)

The purpose of any or all of these forms of governmental participation in student lending is to provide ways for eligible students to defer their share of the higher education cost burden to the future, thus supplementing governmental revenue with some revenue from the student (and maintaining, or even expanding, higher educational participation) at less cost to the government, or general taxpayer.

II
Forms of Student Lending

Student lending may take several very different forms, but all forms have in common the assumption of some of the costs of higher education—either costs of instruction (i.e. tuition), or other educational costs (such as books and supplies), or the costs of student living (such as room, board, and other expenses)—by the student, but deferred into the future. Thus, some lender (whether the university, the government, a bank, or any other saver) is bearing these costs in the present, but will get repaid with interest, by the student (perhaps with the government’s help) as borrower.

Student loans may take one of two basic forms, with many variations of each and with "hybrids" of the two also possible.

Conventional Loans. A conventional, or "mortgage-type," loan carries a rate of interest expressed as an annual percentage of the amount borrowed, a repayment period, or the amount of time the borrower has to repay the loan, and repayment terms, such as whether the payments are to be in equal monthly installments, or installments that begin small and increase over time, or some other arrangement that yields a stream of payments sufficient to amortize the loan at the contractual rate of interest.

Income Contingent Loans. An income contingent (or "contingent repayment") loan carries a contractual obligation to repay some percentage of future earnings (sometimes per $1000 borrowed), generally until the loan is repaid at a contractual rate of interest, or until the borrower has repaid either a maximum amount (which can release the high earner) or for a maximum number of years (which can ultimately release the low earner). That which is stipulated in the loan contract is the annual repayment burden, or the percentage of earnings that must go to loan repayment (which may be fixed for all income levels, or progressive, increasing at higher incomes). That which varies according to income or earnings, in the simplest form, is the repayment period. This is in contrast to a conventional loan in which the repayment period as well as the rate of interest is stipulated in the loan contract (either fixed for the duration of the repayment period, or pegged to some measure of inflation or cost of money), but the annual burden of repayments—the relationship of repayments to current earnings—varies.

Income contingent loans are frequently, but mistakenly, portrayed as always less costly to those borrowers who turn out to earn a low income. But this is true only of those income contingent loans that have an alternative source of subsidy to make up for a stream of repayments from low income borrowers that can never repay at the full cost of the money (i.e. at the full rate of interest). This source may be the government, which may ultimately make up the shortfalls from the low earners in the same way that it might elect to make up the shortfalls from borrowers who simply default, or might provide other kinds of grants or subsidies to students on the basis of their low family incomes at the time they were in the university. Or, the source of subsidy may be the high-earners who, in this type of plan, would have to finish their repayments having repaid at a premium rate of interest, thus effectively subsidizing their low-earning borrowing colleagues and providing the loan program with an average break even interest rate over all of the loans. (The principal conceptual flaw in this concept—perhaps explaining why there are no such mutualized plans in operation— is that students who reasonably anticipate high lifetime incomes will decline to participate, at least in any voluntary scheme, thus depriving the plan of its necessary source of subsidies to protect the low earners.). However, an income contingent loan program can also feature very little governmental subsidization, and no premium rates from high earners (i.e. who merely complete their repayments more quickly), thus requiring that low earners also repay their loans at the stipulated interest rate—but require very long repayment periods to do so.

Graduate Tax. A variant on the income contingent loan is the graduate tax, whereby the student (sometimes only the graduated student), in return for government subsidization of higher education in the form of low or no tuition (and possible of an additional student maintenance grant), becomes obligated to an income surtax, generally for the rest of his or her earning lifetime. A true graduate tax is just that: an income surtax on university graduates, without the keeping of individual borrower accounts or "balances owed." However, one purpose of a graduate tax—like any governmentally-sponsored student loan plan--is to shift a portion of the costs of higher education from the government or taxpayers to students, but to be paid only after the student has finished (presumably graduated) and is earning an income (supposedly higher because of the higher educational experience). The financial success of the graduate tax is measured by the discounted present value of this stream of future income surtax payments—just as the financial success of a government-sponsored income contingent student loan plan, is measured by the present discounted value of repayments that are based on a percentage of yearly income. Thus the mathematics and the practical effect on participating students of the graduate tax and the income contingent loan—assuming similar terms—are practically indistinguishable.

The principal conceptual flaw in the graduate tax concept—at least when taken literally as a tax rather than a loan--is that taxes and surtaxes, unlike contracts to repay, may be abandoned tomorrow by another government. This seriously diminishes the marketable present value of those future surtaxes, making it impossible to tap a private capital market and forcing governments to treat the initial outlay of funds, whether to students or to the universities, as simply another governmental expenditure—which is precisely what the concept of student loans is attempting to lessen in the first place. This may explain why, although several countries have operational, governmentally-sponsored income contingent student loan programs (Sweden, Australia, New Zealand, Australia, UK, and the US), there has to date been no successful operational scheme of a true graduate tax.

III
Examples of National Student Loan Programs

Some national student loan programs, as of 1999-2000, provide examples of the conceptual models described above.

The United States: provides mainly conventional loans, generally available to all stu-dents with some financial need (including some students from upper-middle income families attending very expensive private colleges and universities), at minimally-subsidized rates of interest, with the federal government subsidizing the interest during the "in-school" years for needy students, and bearing most of the risk of default for all students. Much of the capital and loan origination is provided by the private banking sector, although the federal government can lend to students directly, in turn either selling the notes in the private capital market or tapping the federal government’s general borrowing capacity. Student borrowers can elect to repay according to an income contingent repayment schedule, but as yet relatively few have elected this repayment option, which features mainly convenience and little ultimate low-earnings protection.

Sweden: Sweden (along with other Scandinavian countries) has relied on student loan programs since the 1960s to cover student living costs and to free parents from the obligations of paying for these costs. (The university is tuition-free; that is, the government already pays all instructional costs.) Swedish student loans are generally available—that is, available to all who wish to avail themselves of the opportunity, with no "risk rating" or co-signatory requirement, and diminished only according to the students own income and/or assets. Loans carry a nominal interest rate that mirrors the rising cost of living—that is, a zero real rate of interest. Repayment is income contingent at a flat (i.e. non-progressive) rate of 4 percent of income, with most students ultimately repaying at the effective rate of inflation, and with very low defaults.

Germany: Germany has an extensive system of means-tested, or "need-based," study assistance known colloquially as BaföG. At different times, different portions of the accumu-lated BaföG grant has been treated as a full grant, and the other portion as "repayable"—i.e. as a loan. In 1999-2000, one-half of the total accumulated study assistance must be repaid—with the first repayment due five years after graduation, at a zero nominal rate of interest (which is actually a negative real rate of interest), and many additional provisions for deferment or forgiveness--making even this supposedly repayable portion mainly an effective (additional) grant, and very little "true loan."

The Netherlands: Student loans are provided in the Netherlands to cover tuition and maintenance. Part of the loan, including a basic allowance that is not "means-tested," plus another means-tested component, can be converted to a grant if satisfactory academic progress is maintained. Interest on the remainder varies annually, at the government's borrowing rate plus about 1 percent to cover administrative costs. Repayments are fixed after a two-year "grace period," with an income contingent payment feature for those whose incomes are low. Repay-ments remaining for those repaying on an income contingent basis are forgiven after 15 years.

United Kingdom: The UK student loan program began in 1989-90 as a small, conven-tional (i.e. mortgage type), strictly "top up" loan program as the government began to freeze, and
even lower, some of the once generous means-tested maintenance grants. The private sector never embraced the program, however, and in 1998-99, a much expanded program was announced by the government to replace the former maintenance grants and to accommodate the inauguration of means-tested tuition. Some borrowing is generally available to most students, with maximum loans means-tested. The As in Sweden, loans carry an interest rate that mirrors the rate of inflation—that is a zero real rate of interest. Loans are administered by the government’s Student Loans Company. Repayment is income contingent--9 percent on marginal income above £10,000 [$14,858] , deducted by the employer as though an income surtax and passed to the government treasury. Any indebtedness remaining after the borrower reaches age 65 is to be forgiven.

Australia: The Higher Education Contribution Scheme (HECS) in Australia is a generally available loan program up to the full amount of tuition, which in the year 2000 ranges in three "bands" from $A 3463 [$2664] for arts and sciences, nursing, and education), to $A 5772 [$4440] (for law and medicine). Up to 25 percent of the tuition due is discounted for paying "up front." The interest rate, as in Sweden and the UK, mirrors the rate of inflation—that is a zero real rate of interest. Repayments are income contingent on annual incomes above A$ 21,984 [$16,910]. Rates range from 3 percent to a maximum of 6 percent on annual incomes in excess of A$ 39,573 [$30,440]. Repayments due are collected as an 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 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. (Significantly, both of these results could be attained as well with a conventional student loan plan that similarly enlisted the authority and efficiency of the national tax system, or otherwise mandated employer withholding for student borrowers in repayment.)

South Africa: Student loans are given by the governmentally-sponsored Tertiary Education Fund of South Africa (TEFSA). Loan amounts range between R 1100 and 13,000 [$200-2,364] , and are need-based. The interest rate is a relatively high inflation-plus-two-percentage-points, with no in-school interest subsidy. However, fully 40 percent of the amount borrowed can be converted to a grant if all subjects are passed, with this "forgiveness" prorated for only some subjects passed. Repayment is income contingent, beginning with 3 percent on the first R 26,300 [$4782] of income, progressively adding an additional 1 percent for each annual income increment of R 6000 [$1090] until a maximum of 8 percent of income must be paid for student debt retirement at an annual income of R59,300 [$10,782] and above. The national tax and pension contribution systems are not used for collection, but the government has authorized the tax agency to report borrower incomes to TEFSA for purposes of income verification.

Kenya: Kenya began its current Higher Education Loan Scheme under a quasi-public Higher Education Board in 1995. Amounts are "means-tested." Interest is 4 percent. The key to the program is the extraordinary effort to recover payments, based on the new legislation that mandates employers to deduct amounts due from employees in repayment. (The new law also mandates employers to collect outstanding loans from the earlier, 1974, Kenyan student loan program.) Non-payment is supposed to bring severe penalties to both borrower and employer. Capital is provided by the government, with the hope that repayment on the "old loans," buttressed by the vigorous collections on the new loans, will soon capitalize a revolving student loan fund.

China: China announced an experimental demonstration loan program in the summer of 1999 in eight cities. Loans were not, however, guaranteed by the government, and parental or other family co-signatories and pledged collateral were necessary for most student borrowers. Thus, the loans were not available for the truly financially needy except for those at institutions that would assume all risk of non-repayment—which none elected to do. The interest rate was the prevailing commercial rate (10.8 percent in the spring of 2000), with one-half paid by the government and one-half by the borrower. However, the repayment periods were extremely short (four years), placing a high annual repayment burden on the borrowers. As of the 2000-2001 academic year, the program is still "in transition."

The brief descriptions above are provided simply to illustrate some of the dimensions of variation suggested in the sections above. The next section returns to simplified models of generic student loan programs.

IV
Elements of a Student Loan Program

The elements that fully describe a student loan program, including the repayment obliga-tion, are the following. They are summarized here not to present a "cookbook" for the start-up of a student loan program, but to illustrate the key elements, possible "players," variations and complexities in student lending.

1. The Governmental Authority (if any): the underlying law, regulations, and administering governmental agencies underlying the government’s sponsorship of student lending. Absent a governmental sponsorship, student lending would be presumed to be private, under the sponsorship or ownership of either a bank, a non-governmental organization, or a higher educational institution, and subject to the limitations discussed in Section I (i.e. of availability restricted to children of creditworthy parents, and possibly with excessive interest rates).

2. Eligibility: who is eligible to borrow? Are there limiting criteria such as full-time student status, study only in certain institutions, or in certain academic programs? Of special importance are criteria of "financial need" (usually determined according to the financial means of the parents), or "academic merit" (which may refer either to academic promise or actual performance).
3. Terms and Conditions of the Loans. The critical terms and conditions of a government-ally-sponsored student loan program—whether conventional or income contingent—are the following:

a) Origination: How, and from what institution or agency, are the loans actually disbursed, and with what entity does the student borrower and any required co-signatories make a legally enforceable contract? (This may be a governmental agency, a quasi-governmental "public corporation," a private bank, or the higher educational institution itself.) In some cases (e.g. Germany or South Africa), the loan is that ultimately repayable part of a larger sum given to the student as "study assistance"--the other part being a grant, or bursary. For loans given to students at public institutions and which are limited to no more than the tuition due, no cash need actually change hands, and the "loan" (as in Australia) becomes whatever amount of the cost of higher education the student is to bear (i.e. the tuition) and then chooses to defer rather than to pay "up front."

b) Loan Amounts and Limits: How much can be borrowed (or deferred), yearly and in the aggregate? To significantly enhance accessibility (and not merely provide a better standard of student living, or reduce the amount that might otherwise be contributed by the parents), the maximum loan should be sufficient to cover at least the minimum expenses associated with university participation, less any reasonably expected means-tested contribution from parents, and less any amount possible and deemed appropriate for the student to earn and save during the term or between academic years. At the same time, the resulting aggregate debt levels—together with the interest rates and repayment periods that together generate the monthly repayment amounts--must be in some kind of accord with the prevailing earning of the graduates so that repayment is possible without great hardship (and thus likely default).

c) Rate of Interest: Money, in whatever currency, has value and must be paid for. In addition to the cost of money itself (which will be some rate of interest in excess of the prevailing rate of inflation for there to be any real return to the lender), the costs of a loan program must include the costs of defaults and the costs of loan servicing. The key question in student lending is how much of this cost is to be paid for by the student borrower in interest, and how much by some source of subsidy—generally, as covered above, by the government, or taxpayer? (Some student loan schemes, such as those in Sweden, Australia, and the UK, will disclaim that they charge "interest," as such, and purport merely that they adjust upwards the amount owed according to the prevailing rate of inflation. This assures that the borrower repays in real terms only what he or she borrowed. But this is still an interest rate, albeit one that is clearly subsidized and that cannot be determined until the loan has been fully repaid.)

Nearly all student loan programs—and certainly all generally available loan programs--require a subsidy. A government (i.e. taxpayer) guarantee is one such very real, although indirect, form of public subsidy, in the absence of which student loans, without a parental co-signatory, would be prohibitively expensive, if available at all. But even with a parental co-signatory, or even in addition to a governmental guarantee, student loans would be expensive—i.e. carrying considerably higher rates of interest than found on ordinary business or consumer debt--because of their generally small size and high servicing and collection costs.

Perhaps the most critical single term in a student loan program, then, is the rate of interest paid by the student borrower, calculated from the origination of the loan to the point of full repayment. The subsidies, in addition to the very considerable implicit subsidy of any governmental guarantee, would then be represented by a stream of differences between what the student pays each year (which may be nothing during the "in-school" years) and the amount he or she would have had to pay to cover realistically the cost of money plus the costs of servicing and collections. If the lender is the government, the subsidy may not be identified as such, but can be roughly calculated by the amount originally borrowed minus the present discounted value of the expected repayment stream. If the lender is private--such as a bank, an NGO, or a university--the subsidies (again, in addition to the implicit subsidy of any government guarantee) would have to be explicit. Possible representative interest rates on generally secure (i.e. with a government guarantee or a parental co-signatory) student loans are shown in Table 1.

Table 1
Representative Interest Rates on Student Lending

Interest Rate a
Explanation Country
Examples

High Market Similar to consumer debt generally. Private plans in any country, without gov-ernmental participation.

Low Market At or near the government’s borrowing, or the prime commercial, rate. US (unsubsidized) Netherlands, South Africa, Japan Type II

Zero Real Interest rate mirroring rate of inflation (somewhat below "low market"). Sweden, Australia, UK
Moderately Subsidized Low interest; substantially below market. US (subsidized)
Kenya

Highly Subsidized
Very low or No interest rate. Germany
Japan Type I
a As calculated from origin of loan to final payment.

d) Repayment Period: What is the repayment period, and does this period vary with the aggregate amount borrowed, allowing more time to repay larger accumulated debts? With subsidized loans, the value of the subsidy to the borrower--likewise, the cost to the government lender--increases as the repayment period increases. With income contingent loans, the repayment period will be a function of repayment amounts and thus of incomes, but there may still, as in Sweden and the UK, be a maximum repayment period beyond which remaining debts are forgiven.

e) Provisions (if any) for Deferment, Forbearance, or Forgiveness? May repayments be either refinanced or forgiven in the event of unemployment or other forms of financial hardship? For income contingent loans, there may be a limit, as suggested above, on the number of years, or the age of the borrower, that repayments have to be made, thus providing some ultimate subsidy for "lifetime low earners." For "hybrid" conventional-income contingent loans, with a maximum percent of income that is to go toward repayment of student debt, amounts owed in excess of this percentage may actually be forgiven, or may merely be deferred, or "refinanced."

f) The "Shape" or Nature of the Repayment Obligation: For conventional loans, are the repayments to be monthly, quarterly, or annual? Are they to be in equal installments, or are they (or can they be) graduated upward over time to correspond better with the likely increases in income or earnings? (That is, can they be made to approximate "income contingent" payments, but on a conventional fixed schedule?) Are the repayments to be paid by the borrower, or can they be (or, as in most income contingent schemes, must they be) removed from the borrower’s pay by the employer or the government, similar to the withholding of taxes or pension contributions?

g) For Income Contingent Loans, the Repayment Rate: what is the percentage of income that is required for repayment, and how is "income" to be defined? Is there an "income threshold" that must be exceeded before the repayment rate, or surtax, takes effect, or an "income exemption," such that only income in excess of this amount is subject to the repayment rate or surtax?

h) For Income Contingent Loans, the Repayment period: For how long is this percentage of income to be paid e.g. until the loan is fully repaid at some (what?) rate of interest, or until the attainment of some age, or until some maximum number of years has passed since the beginning of repayments?

4. Bearing Risk, or Guaranteeing Repayment. The concept of risk, as outlined in Section I, is all-important in lending. What is special about a generally available student loan program (that is, one without a test of lender or parental credit-worthiness) is the fact that neither the average student in need of a loan, nor many his or her parents, are apt to have sufficient assets to cover the loan. This imposes a substantial element of risk on the lender and therefore calls for the government either to be the direct lender or to provide guarantees of repayment to a private lender. Risk in more selective student loan programs can be significantly less and may be assumed by the lender—which may be the government itself, a bank, a non-profit or a quasi-public agency. In these cases, even with less risk, the lender may still attempt to protect itself either through the requirement of a co-signatory (usually the parents, who may have to pledge collateral), or through a bad debt reserve, built up from the surpluses of premium interest rates. If the university itself, however, or some other non-profit agency is the lender and bearer-of-risk, it is unlikely to be able, financially, to lend or to guarantee more than a small number of student loans.

5. Providing the Capital. If the loans are secured by a creditworthy guarantor (which in most highly developed countries could be the government, but in some countries might have to be an individual or a corporation with liquid assets or some other collateral), private, and even global, capital markets can be tapped, generally at favorable rates of interest. In the absence of a guarantee, all of the capital will probably continue to have to be provided by the government. This may be a significant limiting factor, as a major underlying purpose of student loan programs is the presumed inability of the government—especially in low-income, or developing, or "transitional" countries--to provide all of the necessary revenues to support both the institutions of higher education and the expenses of student living.

There may also be an initial, and a separate longer-range, source of capital, with the initial source providing the capital at origination, but then either selling or "warehousing" the student loan notes to a more primary source of capital. This is the case in the US, and can be in other countries with extensive and sophisticated financial institutions. (In the US, for example, banks provide most of the initial capital, but then frequently sell the loan notes, which are guaranteed and which bear a near-market rate of interest, in the secondary capital market.) Even in such countries, however, student loan notes that are unfamiliar or may otherwise be deemed risky by private capital sources—such as income contingent loan notes or graduate tax obligations—may not find a buyer in the private capital markets, and may thus may continue to depend on all governmental, or taxpayer, financing. (This is thus far the condition of those loan plans that are income contingent, such as the Australian HECS) plan, which may be quite financially viable, but which must nevertheless depend on tax revenues or on the capacity of the government to provide all of the loan capital.)

V
The Functions and Agents of Student Lending

What the above-mentioned "elements of student lending" suggest is that there are several distinct and quite separable functions to the process of making loans to students, just as there are several different agents or parties that may perform these functions. For example, Sections I and IV suggested that there can be no generally available student loan program without governmental participation. However, this does not mean that there cannot be significant private participation in governmentally sponsored student lending, including but not limited to the private banking system. Nor does it mean that the loan program cannot provide substantial cost recovery from the students. In fact, student lending involves five quite distinct functions, each of which has one or more appropriate agents, only some of which need to be governmental or public. The effective analysis of operating student loan programs, as well as the planning of new or the modification of existing programs, requires the disaggregation of student lending into these several conceptually distinct functions, namely: (1) originating the loan; (2) providing the capital; (3) guaranteeing, or bearing the risk of default; (4) subsidizing some of the cost of the borrowing; and (5) servicing and collecting (including the difficult and expensive task of collecting loans that are in arrears).

These functions are to be found in some form in all student loan programs in any country or situation. In turn, there are alternative agents, or parties, that may be more or less appropriate to these several functions, but that will vary depending on the particular time, culture, country, or purpose to be served by the student loan program. These agents, together with some possible functions, include:

• government: as guarantor, or bearer of risk, as subsidizer, and as possible provider of capital and servicer, or collector;
  
  
• private banks: as originator, capital provider, or servicer and collector;
  
  
• other private sources of capital (such as pension funds, investment banks, or other sources of global capital): as possible purchaser of already-originated loan notes;
  
  
• government ministry or agency (both central and state, or provincial): as regulator and possibly as originator, servicer, and/or collector;
  
  
• quasi-public loan bank: as an originator of student loans, or as an intermediary between the borrower and/or the originator, and the provider of capital; /or pension specialized private loan servicing and collecting agency: for servicing and collecting;
  
  
• international development agencies (e.g. the World Bank): for "start up" capital or risk bearing and partial subsidization only for demonstration projects;
  
  
• universities and other institutions of higher education: as originators, possibly as guarantors, and possibly (but not desirably) as services/collectors;
  
  
• parents: as guarantor, or bearer of risk (by co-signing);
  
  
• community or non-governmental organization: as guarantor or bearer of risk (by co-signing).
  
  

These functions with the most likely agents are shown in Table 2. This interplay of functions and agents, for example, illustrates the two functions, first mentioned in Section I, that most require the participation of government: (1) guaranteeing repayment (at least if the student loans are to be made generally available); and (2) subsidizing interest (if the interest rate to the student borrower is to be less than the rate for general, medium-term, unsecured, consumer debt). Table 2 also illustrates that a number of agents can provide servicing and collection. The most effective, at least in some countries, is likely to be the state treasury, or whatever government agency collects income taxes and/or pension contributions. This is because it is likely to have most citizens and most employers within its system, and to be able to collect student loan repayments along with the collection or withholding of income taxes and /or pension contributions at the point of payment of wages and salaries. This is the case with the Higher Education Contribution Scheme in Australia, and is hoped to be the case with the new student loan program in the UK.

However, there are also reasons to not involve the government’s tax collecting authority and apparatus in the collection of student loans. The belief within the US government is that maintenance of the much-vaunted US voluntary income tax compliance is so important for the preservation of the US public revenue base that anything that might compromise this compliance (such as servicing student loans) is almost certainly not worth whatever advantages it might provide in the cost-effectiveness of loan collections. There might still be a law providing for the voluntary (or even the mandatory) collection of federally sponsored student loan repayments at the employment site--much like the withholding of taxes. But the involvement of the US Internal Revenue Service in anything extraneous to the collection of lawfully owed taxes has been resisted.

Furthermore, many developing countries have notoriously inefficient and ineffective income tax collection systems already, which the added burden of student loan collections might simply aggravate. In addition, the collection of amounts due the government via the tax system looks very much like the earmarking of tax revenues for a specific purpose—and there is no end
Table 2
Functions and Agents of Student Lending

Function Alternative
Agents
Notes
1. Origination: Loans have to be put into the hands of the student (or given directly to the institutions on behalf of the student), contractually: i.e. in the proper amounts and on the proper repayment terms. &Mac183; Private banks
-Universities themselves
-Public or quasi-public student loan entities Origination is essentially paperwork and compliance with applicable regulations, including verification of attendance and "financial need" or "parental means."
2. Provision of Capital: Capital has to be provided: from savers, or lenders, who are collectively selling (renting) their claims on goods and services to the buyers, or rent-ers, or borrowers. The large sources of primary capital in industrialized countries are public and private pension funds, insurance companies, mutual funds, banks, etc. &Mac183; Primary capital sources
-Global capital sources
-Private banks
-Government (taxpayers)
-Government (deficits & currency inflation) Banks are important, but not essential if other more primary capital sources can be tapped—which is not difficult with guaranteed assets. The public treasury (tax revenues) should be last resort—turned to only when primary capital is unavailable due to perceived risk.
3. Bearing the Risk: The fundamental problem of lending to students is risk, as they generally have neither a credit rating nor assets to use as collateral. Neither banks nor universities can be expected to provide loans without some other entity assuming the risk—i.e., guaranteeing full or nearly full repayment. &Mac183; Government
-Parents & other co-signatories
-Possibly:
---Universities [?]
---Internat. Dev. Agencies [?]
---Other Borrowers [?] If a student loan program is to be generally available –i.e. regardless of creditworthiness of parents—the principal bearer of risk must be the government. Others that can share in the risk are parents or other entities that are charged a premi- um interest to spread the risk among borrowers.
4. Subsidization (in addition to the extensive Subsidization implied by a guarantee): Subsidies lower the cost to student borrowers (or to some borrowers) below the actual cost of the capital. In the absence of some subsidies—and even assuming guarantees--student loans could only break even at interest rates similar (at best) to other consumer loans.
-Government Subsidization of student loans can enhance the attractiveness (to the student) of borrowing generally. Or, it can selectively subsidize certain borrower behavior: e.g. good academic performance, prompt repayment, or decisions to go on the post graduate study or to practice a certain occupation or in a certain location.
5. Servicing and Collection: this function includes keeping track of repayments, making the special efforts to collect loans in arrears, and ultimately collecting from guarantors or co-signatories. &Mac183; Banks
-Private loan servicing entities
-public student loan agencies
-Employers via withholding
-Government via tax withholding Loan servicing is a business that can, but does not need to, be associated with banks or other financial institutions. It mainly requires telephones, computers, and labor.
to the claimants and purposes that would like to enlist the government’s tax collector in their cause. Thus, while collection of student loan repayments via the tax system has a certain appeal, it also has limitations and possible "downsides."

The array of functions and agents shown in Table 2 also suggests some functions that may be most desirably privatized. One such function—and the one that may be the most difficult in low-income countries—is the provision of the loan capital itself. Student loan notes are assets that have at least potential value in the private capital market. Student loans that cannot either be made by, or ever sold to, private capital sources must forever depend on public funds, and must consequently displace other uses of public revenue such as elementary and secondary education, public health, housing, public infrastructure, and income maintenance. Therefore, it is desirable for the student loan capital to come as much as possible from savers directly or from their agents, such as banks (that is, from the private capital market), rather than from governments or taxpayers.

The ability of a student loan plan to tap private capital sources depends on the value of the student loan notes themselves. This value, in turn, depends first on the certainty of repayment (which depends on the perceived creditworthiness of the borrower or the guarantor), and second on the rate of interest on the notes, which together (less any subsidies) yields a present discounted value of the anticipated repayments. In this light, excessive governmental subsidization of interest is not only a direct cost to the government, but also lowers the worth of the student loan notes in the marketplace and lessens the likelihood that a private capital source will be found to relieve the government from the burden of providing all of the capital for the student loan program, in addition to bearing the direct of the subsidy and the contingent costs of the guarantee.

V
The Purpose of Student Lending

The purpose of student lending, with special reference to developing countries, has been deferred to this section so that the purposes and benefits can be seen in the context of the rich variability and financial complexity of student loan programs, described in the preceding sections. There is a special case, or need, for student loan programs in developing countries. Yet there are also very special problems, many of which have been outlined in reports to the World Bank and the International Institute for Educational Planning by Woodhall in the early 1990s, and by Ziderman and Albrecht in 1995. Many student loan programs in developing countries have disappeared; most that remain are small and selective. Few if any have reported substantial cost recovery. Yet the case remains, and new student loan programs appear regularly in the developing would. Jamil Salmi, drawing on World Bank involvement, reported new or reformed loan programs in Jamaica in 1996; Hungary, Mexico, and Malaysia in 1998; and programs in preparation or under study in Namibia, Ethiopia, Brazil, and Lebanon in 1999.

Cost sharing. The case for student lending in developing countries, as in all countries, is derived from the case for higher educational cost sharing. Cost sharing simply means that parents and/or students share with government, or taxpayers, the cost of higher education. In turn, this suggests both tuition (in the so-called "public," as well as the "private," sectors of higher education), and at least some parental and/or student assumption of the costs of student living—i.e. lodging, food, and other expenses. Since higher education in very many countries--and especially in developing countries, as well as those countries "in transition" from nearly total state ownership and control of production to more market-oriented economies--have featured either low or no tuition, plus heavily subsidized lodging and food (frequently via once-generous student stipends), cost sharing means a significant and often strongly-resisted shift of the higher educational expense burden from government, or taxpayer, to parent and/or student.

The Equity Rationale. The case for cost-sharing for all countries, but applying especially to most of the so-called developing world, can be made on the basis of either (a) equity, or (b) the diminished relative public priority of higher education. The equity case is the familiar one pressed by generations of market-oriented economists, and is borne of two observations. First, students everywhere, but especially in developing countries, tend to be from the most affluent families, most of whom both could and would pay at least something toward the costs of their children’s higher education. Second, public revenues in most countries--again, especially in developing ones--are raised mainly in proportional (at best) or regressive ways, either from deficit financing and inflation, or from business, consumption, or value-added taxes that are ultimately borne by the general consumer. Hence, all citizens, including the very poorest, are paying for an expensive product (higher education) that is enjoyed largely by an elite, which brings both parents and their children even further advantages that they would be quite willing to pay for (or at least share in the expense burden) if they had to. This assertion is supported by the substantial number of families who do pay for private higher education or higher education abroad when the free public higher education is unavailable or unsuitable for their children. By this line of reasoning, it is more equitable for those who can and will pay to be expected and required to do so—both through some tuition contribution toward the instructional costs of higher education, and through paying full or nearly full costs of any institutionally- or governmentally-provided lodgings and food.

Implicit in the equity case for greater cost sharing is the assumption that it is both appropriate and "fair" to expect contributions from both parents and students. But the rationale and the mechanism for a parental share are somewhat different from the rationale and mechanism for a student share. And it is the rationale for the student share that is most fundamental to the rationale for a student loan program. Most of the parental contributions are expected to come from savings or assets and current income. It is thought both possible and reasonable to expect parental contributions because parents clearly benefit from the satisfaction and greater status of seeing their children succeed; and higher education is increasingly believed to be necessary to life success. Most cultures believe that the parental financial responsibility (to the extent the parents are financially able) extends through "young adulthood," and probably through at least the first higher education degree. Furthermore, we know that there is a benefit—or at least a value--to parents in the higher education of their children by the simple observation that parents do pay (to the extent of their ability) when they have to. In a market economy, that willingness to pay is the predominate signal of value.

The benefits to the students are even more obvious, and extend to the prospect of higher lifetime income, greater status, more "lifetime options," and all of the personal non-monetary satisfactions that come from being more highly educated. These benefits have nothing to do with the income of the parents, or with capturing the "low-hanging fruit" of the willingness of affluent parents’ to share in the higher educational expenses of their children’s higher education. But most students, even of financially well-off parents, have little income of their own. Thus, for the student to share in the expenses of his or her higher education virtually requires some sort of loan, or "deferred payment," program.

The Increased Need (for Non-Governmental Revenue) Rationale. A quite different rationale for cost sharing--that is, a greater parental and/or student contribution to the costs of higher education--is the greatly increased need for non-governmental revenue, including from parents and/or students. The increased need rationale is quite apart from any theoretically greater equity to parents and students assuming a larger share of higher educational costs (although, like the equity argument, it is probably even more compelling in the developing world than in industrialized, highly-developed countries). The increased need, argument, rather, rests on three observations.

First, the demand for costly higher education is being driven upward very rapidly by the combination of high birth rates, the increasing proportions of these larger age cohorts who are completing secondary education and who are thus being at least theoretically prepared for higher education, and finally the increased demand (for all of the obvious reasons) of these exploding numbers for some higher education.

The second observation is that public revenues—that is, the taxing capacity--of most countries, and especially of developing countries (with the exception of those lucky few whose public revenues are derived mainly from petroleum production) are extremely limited. The shrinking of state-owned productive enterprises has diminished the effectiveness of the value-added taxes in the former Socialist economies. Globalization has made it too easy for wealthy individuals as well as production facilities to escape to lower tax jurisdictions. And the globalization of capital markets and free-floating currencies has made it just as difficult to raise public revenue through deficit financing and inflation.

The third observation is that other competing claims for this increasingly limited public revenue are also growing, and often enjoy greater political popularity. In most countries, primary education, transportation, telecommunications and other public infra-structure, public health and sanitation, and income maintenance (to name just a few) are frequently ahead of higher education in the queue for the available tax dollar.

Thus, the need for other-than-governmental revenue to support the growing financial needs of higher education is huge. And regardless of how convinced one may or may not be by the classical neo-liberal economic equity justification for cost-sharing, countries may be forced to shift at least some costs from government to either parents (those who can pay) or students, or both by the sheer unavailability of alternative revenue. And while the principal target may be well-to-do parents who can and will pay, it is likely that there will continue to be a search for a workable student loan program, as well, to supplement the increasingly limited public revenue.

VI
The Imperative of Cost Recovery

For either rationale underlying a case for student lending, the effectiveness of a student loan program depends on the actual cost recovery. And the problem with student loan programs, especially in developing countries, is that there are significant, and exceedingly hard-to-stem "leaks," or sources of loss, in the cost recovery of most student loan programs. The first of these "leaks" is defaults. Default data is inconsistent and unreliable, especially in the small and unstable loan programs characteristic of most developing countries, but Ziderman and Albrecht reported defaults from loan programs in the 1980s in Brazil, Venezuela, and Kenya estimated at more than 90 percent. And if these estimates are off by a factor of two or even three, they still represent a significant erosion of potential cost recovery.

High default rates are almost to be expected from student loans absent a culture that accepts the appropriateness of borrowing for higher education and absent systematic and vigorous efforts to secure repayment. A reasonable repayment (i.e. low default) rate depends on: (a) the clear understanding on the part of the student that the initial stipend was a loan to be repaid, and that there would be consequences for default; (b) good record keeping on the part of the university and the lender so that repayment begins in timely fashion upon graduation or other departure from study; (c) an economy that holds forth the prospect of most graduates finding employment, making repayment possible; (d) a culture that accepts the very notion of cost sharing, and therefore the appropriateness of borrowing for higher education—and thus the importance of repayment; and finally (e) effective and efficient collection, both from student borrowers and from any co-signatories, on loans in arrears or default. Most of these attributes are in short supply in the developing and so-called "transitional" countries.

The second source of loss or "leakage" from the repayment stream is excessive governmental subsidization. All student loans that bear less than full market rates of interest are, in effect, a combination of:

&Mac183; a true loan, or that portion of the original initial amount borrowed that the subsidized repayments actually amortize (or expressed another way, the dis-counted present value of the actual stream of repayments); and

&Mac183; an effective grant, which can be viewed as the present value of the stream of loan subsidies, or the difference between the original amount borrowed and the true loan amount. /

Loans may be excessively subsidized because of the political volatility of cost sharing in any form and the economically irrational, but politically understandable, notion that students or their parents will be mollified if the loan bears only a low (and therefore highly subsidized) rate of interest. Loans may also be highly subsidized because of other public policy objectives that the government may elect to pursue through targeted subsidies via repayment forgiveness. For example, the US government has subsidized loans through a forgiveness of a portion of the amount owed for each year that the graduate (now in the loan repayment stage) pursues an occupation in a career (e.g. teaching), or in a region (e.g. difficult "inner city" schools), that the government is trying to encourage. Germany timely degree completion with additional repayment forgiveness. South Africa uses forgiveness of loan repayments to reward successful completion of the academic program. These examples may—or may not—be cost-effective ways to induce the desired behavior sought by the government policy. But successful cost recovery of the student loan program needs to be judged independently of these ancillary features, which are really government expenditures in pursuit of a public policy objective having nothing to do with the underlying rationale of cost sharing, or the shift of some higher education costs from the government or taxpayer to the student.

Finally, the effective loan recovery depends on the costs of administration and servicing the loans. Again, good data are sketchy. Ziderman and Albrecht estimated 2 percent or less in most countries. But the general presumption is that administration and collection costs, especially in developing countries, are very high. Part of the reason is the relatively small size of the average loan, discouraging most applications of economies of scale. Another reason is the mobility of recent university graduates, which makes the tracking of students in repayment both difficult and expensive. Finally, the servicing of student loans in developing (and most other) countries has done by government agencies, with all of the inefficiencies so frequently associated with government bureaucracies and the civil service. And although a number of highly-developed countries such as Australia, Sweden, and the UK--with effective and efficient tax collection systems--have elected to collect student loan repayments through these agencies, this may not yet be an option in most developing and "transitional;" countries.

Combining the "leakage" from defaults, subsidization, and administrative costs, Ziderman and Albrecht presented an array of estimated "effective cost recoveries" on various student loan programs in ranging from outright losses (i.e. where the present value of repayments, diminished by subsidies and defaults, fails even to cover administrative costs), to the long-established and generally effective income-contingent loan programs in Sweden, which they estimated to recover 67 percent of the amount lent. The record of student loan programs in countries that may be termed "developing," or "less industrialized," or "transitional" is not encouraging—both from the lack of substantial real cost recovery, and also from the continuing political unpopularity of such manifestations of cost sharing as public sector tuition or higher educational privatization.

VII
Problems and Limitations of Student Lending

Much of the record of failure alluded to above is due to problems and limitations—some of which are peculiar to low income, or less industrialized, countries, others to the so-called "transitional" countries, and some found in both--that seem, at least in the short run, to be intractable: that is, not solvable simply with "smarter policies" or "better execution." Other problems and limitations may be amenable to alternative policies and, or practices. Among the most serious problems and limitations are the following six:

1. The underlying extremely high per-student cost of higher education, particularly at the university level and particularly relative to the underlying per-capita income of most "less industrialized," or "developing," countries, exacerbated by pressures to greatly expand these already costly enrollments. This is not, strictly speaking, a problem of either cost sharing or of student loan programs per se. But it makes the purpose of cost sharing and lending—i.e. a partial solution to the dilemma of grossly insufficient public revenues for the higher education sector—so daunting and painful that a certain amount of unpopularity and failure is almost inevitable, even with a well-run loan program.

2. The high rates of unemployment and low-paying jobs in some countries among university graduates, making student loan repayments difficult (even with an otherwise well-conceived and well-administered loan program). It is to be hoped that this condition is both temporary and at least partially self-correcting: temporary because many such economies have yet to create substantial numbers of middle- and upper-management and professional jobs, but should do so in the future; and self-correcting, because increasing competition for students and tuition dollars should lead universities to begin to offer the kinds of programs that lead to employment (including what used to be termed short-cycle, non-university programs).

3. The pervasive belief that higher education is or ought to be a public entitlement: that is, paid for by everyone even if only relatively few participate and benefit.

Given the past dominance in so many developing countries of various kinds of socialist ideologies and political-economic systems, a belief in entitlements (especially among those who would be so entitled) is very persisting. But as long as the less-industrialized countries are moving toward more private ownership, market principles, and governmental transparency, the realities of public sector costs and trade-offs become inevitable, and so also does the need for some cost sharing in higher education--and thus some form of student borrowing and lending. (Witness China and Vietnam adopting tuition in the late 1990s while still adhering officially to tenets of Marxism.)

The persistence of the principle of no tuition on the European continent, including Capitalist, Socialist, and Soviet-style centrally-planned systems alike, reinforced this belief in entitlements. However, the European tradition was only for no tuition. The costs of student living, on the other hand, could be quite high, and were assumed to be the responsibility of the family--except in the Nordic countries, where such costs were to be borne principally by the student in the form of loans. At the start of the 21st century, the principle of cost-sharing—and the consequent interest in student loan programs—has been given further impetus in Europe by the advent of tuition in countries like the Netherlands, Portugal, and Britain, all of which adopted tuition in the 1990s, the announcement by the Austrian government in November 2000 of its intention to require tuition, and signals that tuition may not be so far away in Germany. Meanwhile, in Russia and Eastern and Central Europe, free higher education is still constitutionally protected—but only for the most able students. In these countries, the number of fee-paying students within the public sector, in addition to those in the tuition-dependent, non-state sector, continues to grow: an increasing de facto cost-sharing that has caused these countries also to look for relief in student loan programs.

4. The pervasive absence of trust, especially among the student and potential student populations in many countries, of government and of the university administrations. This is hardly "intractable," but it does seem enduring, and it greatly reinforces the political resistance, mentioned above, to "cost sharing." In many countries (Mexico comes to mind, as do many African countries), students and parents who might intellectually accept the case for some cost sharing and student borrowing may continue to resist because they believe that their tuition would go not toward expanding higher educational capacity or quality, but simply toward a corrupt "system" (referring either to the university administration or to the government or to both).

5. The low rate of savings and general scarcity of private capital. The limitation of private capital limits the supply of student loans to whatever the government can make available. This puts student lending in competition with alternative direct governmental outlays, rather than with alternative investments. Or, it limits student lending to that which students and families can borrow themselves in the limited private sector—which may demand the kind of collateral that puts borrowing out of reach of many low-income families.

6. The absence of reliable and cost-effective systems for loan servicing and collecting. Keeping track of loan balances and repayments on relatively small amounts of principal among a very mobile population is expensive—if possible at all—and contributes both to the high default rates and to the high costs of administration and collection. Underlying the problem in most of the less-developed and "transitional" economies is the absence of extensive and efficient banking, postal, and telecommunications services. Even linking loan repayments to employee tax or pension withholding may still be expensive and unreliable—and generally limited to employees of relatively large corporations or the government.

In spite of the very considerable problems and imitations, the stakes in creating a workable student loan program in many countries—i.e. the promise of some real cost recovery from students--makes further efforts worthwhile. At the same time, efforts to date have not only been burdened with these problems and limitations, but also with some persisting myths and "partial truths" about student lending—to which the next section turns.

VIII
The Myths and Partial Truths of Student Lending

The landscape of failed student loan programs, particularly in developing and "transitional" countries, seems (at least in retrospect) to be littered with avoidable myths and misconceptions, or at best "partial truths." Among these "myths and partial truths" of student lending are the following:

1. That student loans (especially in low-income countries) must be heavily subsidized in order to make repayments manageable.

The issue of the manageability of repayments is partly, of course, partly a function of the degree of subsidization as reflected in the effective rate of interest. But equally, or even more, important are: (1) the aggregate amount borrowed and owed at the initiation of repayments; (2) the repayment period; and (3) the "shape" or nature of the repayment obligation—i.e., whether to be repaid in equal installments, or in installments that are graduated over time, or even on an income contingent basis. Students loans must be kept to a manageable total, repayable over a sufficient period to make the actual monthly payments also manageable, and desirably amortized over time in such a way as to recognize that the ability to repay may be minimal in the first years after graduation.

Clearly, a heavily subsidized loan, all else being equal, is less burdensome than a minimally subsidized one. But with the heavily subsidized loan, the government is implicitly accepting a lower cost recovery when, for the same present value of effective governmental expenditure, it could have chosen to require less tuition in the first place. In short, lower tuition with "minimally-subsidized" loans and more modest aggregate indebtedness, repayable over a sufficient number of years and with provisions for handling periods of unemployment or financial hardship, may be quite preferable to large aggregate indebtedness that then needs to be heavily subsidized in order merely to make the repayment burden manageable. A further problem with heavy subsidization is that it mandates very stringent means-tested rationing. Yet one of the problems of any kind of financial assistance in many countries is the great difficult of verifying "means" or "financial need." The less subsidized the loans, the less the government needs to worry about such rationing.

2. That low-income, or rural, or ethnic minority students will not borrow out of a "cultural aversion to debt."

This may be a "partial truth," although it is buttressed more by plausible assertion, conventional wisdom, and repetition than by empirical evidence. Certainly, low-income and minority youth in the US borrow freely. But even if it is at least partially true or true in some countries and some cultures, it ought not dissuade governments from turning to student loan programs to help those students who do wish to forgo some future consumption (i.e. repay a loan) in return for a chance to invest in their higher education (i.e. to borrow). If certain young persons who ought otherwise to be in the university are "culturally averse" to borrowing, they are likely to be similarly "culturally averse" to many features of a modern university education, and public policy will need to address their needs in a comprehensive way, which might entail special grants in lieu of loans. But the initial premise of cost sharing as presented in Section V was that there was, or soon would be, insufficient governmental, or taxpayer, revenue to continue the traditions of free higher education for an expanding student population. If that is the case, then the absence of some effective cost sharing will simply limit higher educational expenditures, and the population to suffer, either from no places at all, or from increasingly shabby and underfunded universities, will be those who have no alternatives--that is, low income, minority, and rural youth. In short, a cultural aversion to borrowing may be a problem to be overcome. But it does not seem to be a reasonable basis for denying student lending altogether.

3. That women will be especially dissuaded from borrowing out of a fear of the "negative dowry" implications of student indebtedness.

Like #2, above, this may well contain a partial truth. But like the above response, the "negative dowry" charge is also in part plausible assertion and conventional wisdom, with little empirical verification or understanding. Furthermore, if some cost-sharing in the form of student indebtedness is nearly inevitable (for reasons given earlier in this paper), and as women also come to participate in the labor market on a more nearly equal basis with men (admittedly far from the case in much of the world), women will need the same educational opportunities, and on the same terms, as men. A less enlightened alternative might be to shield women from indebtedness by directing them lower cost, but also lower prestige, sectors of postsecondary education--as in Japan, for example. A less gender-biased solution would be a combination of reasonable levels of indebtedness for everyone, coupled with manageable repayment terms, provision for periods of financial hardship, and increasing labor market opportunities for women.

4. That income contingent loans are inherently more manageable and therefore inherently "better," at least for the student borrower, than conventional loans, especially for low-income students.

The manageability or burdensomeness of a student loan debt, as covered in the response to "myth" #1, above, is a function of the present discounted value of the repayment burden (the truce cost of the loan), plus the length and "shape" of the repayments. Any debt can seem to have been made less burdensome merely by lengthening the repayment period and lowering the periodic repayments, although this kind of "manageability" is, of course, purchased by greatly increasing the total amount that will have to be repaid over the lifetime of the repayment period. Repayments can also be made more manageable by tailoring them to the borrower’s ability to repay as this ability changes over time—usually in accord with his or her change in earnings. Finally, repayments can be made to seem less burdensome by having them withdrawn by the employer prior to the payment of the wages or salaries. Although most (but not all) income contingent loans schemes have these features, these same features can just as well be built into the repayments of conventional loans. More important, a conventional loan with a lower present value of repayments is still fundamentally less expensive to repay, and therefore arguably "better" for the student borrower, than an income contingent loan with a high interest rate and a high present value of the total expected repayment stream, even if the income contingent loan may seem more "manageable."

5. That income contingent student loans have (inherently) lower default rates and lower administrative costs than conventional student loans.

This is probably at least a "partial truth," although the few operational income contingent student loan programs are either still relatively short-lived (e.g. Australia, New Zealand, and UK) or so deeply imbedded in a culture of acceptance of student borrowing (e.g. Sweden) that it is difficult to say how much of the seeming success is due to the income contingentness of the repayments. To the degree to which either lower defaults or lower servicing costs are due to the involvement of the employer via wage or salary withholding, or to the government via the machinery of tax collection, these same benefits could be achieved with a conventional loan that was collected in the same way.

6. That a student loan program, once funded, or "capitalized," at the outset, or "front end," can in a few years become "self sufficient," with repayments equal to new student lending, and thus never again dependent on either new governmental or new private capital.

This is a recurring theme of new student loan program proposals, with no basis either in theory or in practice. It is extremely unlikely that repayments on past loans will ever be sufficient to become the sole source of new lending, due both to the "leakage" discussed above (i.e. losses from defaults and subsidies), and also to the increasing volume of new lending needed to keep up with the combination of increased dollar needs per student and the increasing numbers of new student borrowers. More important, however, is that there is absolutely no reason even to aspire to a "revolving" or "self-sufficient" loan program. The amounts coming in via repayments have almost nothing to do with the amounts that should be—or can be—lent out to new student borrowers. At its most efficient (and independent of new governmental capital), a student loan program should aspire to having the present discounted value of the anticipated stream of future payment on each cohort of new lending equal to dollar value of those loans. In this way, the volume of student lending can increase (or not) each year, and need have no relationship to the current dollar value of the repayments coming in in that particular year. In short, the notion of a "revolving loan fund" is both unrealistic and unnecessary.

IX
The Future of Student Loan Programs

In light of the fundamental problems and limitations outlined in Section VII, and of the "myths and partial truths" outlined in Section VII, what is a reasonable prognosis for student loan programs in general, and especially in those developing and "transitional" countries that may most need the supplemental revenue for higher educa-tion from expanded cost sharing? The World Bank’s 1994 report, Higher Education: The Lessons of Experience, concluded:

The financial base of public higher education can be strengthened by mobilizing a greater share of the necessary financing from student themselves, who can expect significantly greater lifetime earnings as a result of attending higher education institutions and who often come from families with ample ability to contribute to the costs of higher education. [The report went on to say:] Government-subsidized financial assistance programs for academically qualified yet financially needy students are an essential complement to cost-sharing in higher education…. However, given that in every developing country students attending higher education represent an elite group with income-earning potential significantly higher than that of their peers, it is appropriate that the major form of student financial assistance offered be in government-guaranteed loans rather than grants.

Such a recommendation, however, still does not address the particular problems of student lending in developing or "transitional" countries. Yet from the analysis above, some comfort may be taken by the observation many programs have arguably been flawed, both in concept and in execution. Acknowledging the complexity and the difficulties—both political and technical—of student lending, particularly in other-than highly industrialized countries with well-developed banking and taxation systems, the following recommendations are offered as a conclusion to this study:

1. A policy to move toward greater cost sharing should include a total package addressing:

a) tuition fees,

b) "more nearly full cost recovery" (i.e. a gradual reduction of governmental subsidies) on governmentally- or institutionally-provided lodging and food;

c) some means-tested, or need-based, grant assistance for the most qualified students from the lowest income families,

d) a program of generally available student loans (that is, loans available with regard to the credit-worthiness of the parent or the student’s program of studies).

2. This total package should be well communicated as a long run process that is essential for the expansion of a quality higher education system. The shift of a portion of higher education costs toward parents and students should be accompanied by measures to democratize and liberalize the governance of the campuses as well as the ministry, to effect greater financial transparency, and to share not only the costs but the pain—i.e. to economize on administration and other expenditures not directly supportive of students and of the instructional program. (In short, the student, parents, and faculty need to perceive the possibility of some gain to them—as well as to perceive that the university administration and the ministry have also sacrificed in the name of greater efficiency.)

3. The grants and loans should be sufficient in amount to make possible the enrollment of the most qualified students from the lowest income parents (i.e. parents who cannot bear any part of the expense burden).

4. Grants, including targeted repayment forgiveness or other loan subsidies, should be spent minimally, if at all, on students who would attend and study as effectively in the absence of such grants or targeted subsidies. (In short, minimize the politically popular "academic prizes" and the so-called "merit-based" grants and loan subsidies.)

5. The loans, in accord with recommendation 3, should be of sufficiently long repay-ment periods, and with some provision for accommodating periods of hardship, for the payments to be manageable by most borrowers. (This may entail income contingent repayments or "conventional-but-graduated" repayment obligations.)

6. At the same time, there must be sufficient reasonably-anticipated cost recovery from the repayments that students, faculty, and university leaders can perceive some real financial advantage, either in the form of expanded capacity and/or expanded quality (including some expenditure on other-than personnel). This, in turn, requires:

a) loans being given with the clear understanding of the repayment obligation, and the consequences to students who default;

b) at the very least an interest rate that that recovers the real value (1.e., inflation-adjusted) of the loan; more desirable would be a positive real rate of interest, which would be slightly in excess of the actual or anticipated rate of inflation;

c) repayment guarantees from parents or other sources or sponsors who are financially able to do so and governmental guarantees for those students whose parents cannot;

d) vigorous servicing and collection efforts, possibly enlisting employers through wage withholding and/or the government’s tax collection apparatus, but providing as well for those who may be missed by these "nets."

e) The intention, as least for the future, of obtaining sufficient repayment performance that the student loan notes might have a value in the capital market, and relieve the government of the necessity of 100 percent capitalization, in addition to subsidization and risk-bearing (or risk sharing).

7. All of these recommendations need to be accompanied by continued efforts to gain greater efficiency in the provision of higher education through such familiar recommendations as sector and institutional differentiation, utilization of private higher education, and attention to the potential of increasing economies of scale.

Through such policies and programs, we see some hope that student borrowing, well-conceived and well-executed, could yet play a role in advancing those twin goals, identified in our first paragraph, of "…expanding higher educational participation and accessibility, while simultaneously expanding real cost recovery from students."





References


Albrecht, Douglas, and Adrian Ziderman (1991) "Deferred Cost Recovery for Higher Education." Washington, D.C.: The World Bank.
Bain, Olga (1998) "Cost of Higher Education to Students and Parents in Russia: Tuition Policy Issues." Center for Comparative and Global Studies in Education, University at Buffalo.
Blair, Robert D. D. (1998) "Financing Higher Education in Africa" in UNESCO Regional Office (Dakar Senegal) Higher Education in Africa: Achievements, Challenges and Prospects. Paris: UNESCO.
Johnstone, D. Bruce, (1986) Sharing the Costs of Higher Education: Student Financial Assistance in the United Kingdom, the Federal Republic of Germany, France, Sweden, and the United States. New York: College Entrance Examination Board.
__________ (1992) "Tuition Fees," in Burton R. Clark and Guy Neave, eds., The Encyclopedia of Higher Education, Vol. 2. London: Pergamon Press, pp. 1501-1509.
__________ (1993) "The Costs of Higher Education: Worldwide Issues and Trends for the 1990s" in Altbach, P. G. and Johnstone, D. B., The Funding of Higher Education: International Perspectives. New York: Garland Publishing.
__________ (1999) "Financing Higher Education: Who Should Pay and Other Issues" in Philip G. Altbach, Robert O. Berdahl, and Patricia J. Gumport, Eds. American Higher Education in the 21st Century: Social, Political, and Economic Challenges. Baltimore: The Johns Hopkins University Press.
Johnstone, D. Bruce, Alka Arora, and William Experton, (1998) The Financing and Management of Higher Education: A Status Report on Worldwide Reforms. Washington, DC: The World Bank. [Prepared in conjunction with the October 1998 UNESCO World Conference on Higher Education, Paris, October 5-8, 1998.]
Johnstone, D. Bruce and Preeti Shroff-Mehta (2000) "Higher Education Finance and Accessibility: An International Comparative Examination of Tuition and Financial Assistance Policies." Buffalo: Center for Comparative and Global Studies in Education http://www.gse.buffalo.edu/org/IntHigherEdFinance/. Also, Heather Eggins, ed. (2001 forthcoming) Higher Education Reform. Buckingham UK: Open University Press.
McMahon, W. (1988) "Potential Resource Recovery in Higher Education in the Developing Countries and the Parents’ Expected Contribution," Economics of Education Review, 7 (1).
Osterbeek, Hessel, (1998) "An Economic Analysis of Student Financial Aid Schemes," European Journal of Education. Vol. 33 No. 1, pp. 8-16.
Salmi, Jamil (1999) "Student Loans in an International Perspective: the World Bank Experience." (unpublished).
Stager, David A. A. (1989) Focus on Fees: Alternative Policies for University Tuition Fees. Toronto: Council of Ontario Universities.
US Department of the Treasury and US Department of Education (1995) "A Study of the Feasibility of the IRS Collecting Repayments of Federal Direct Student Loans," June 1995.
Williams, Gareth (1992) Changing Patterns of Finance in Higher Education. Buckingham, England: Society for Research into Higher Education and Open University Press.
Woodhall, Maureen (1988) "Designing a Student Loan Programme for a Developing Country: The Relevance of International Experience." Economics of Education Review, Vol. 7, No. 1:153-161.
_________. (1989) Financial Support for Students: Grants, Loans, or Graduate Tax? London: Kogan Page.
_________. (1990) Student Loans in Higher Education (1) Western Europe and USA. Paris: International Institute for Educational Planning.
_________. (1990) Student Loans in Higher Education (2) Asia. Paris: International Institute for Educational Planning.
_________. (1990) Student Loans in Higher Education (3) English-Speaking Africa. Paris: International Institute for Educational Planning.
_________. (1990) Student Loans in Higher Education (4) Latin America and the Caribbean. Paris: International Institute for Educational Planning.
_________. (1992) "Changing Sources and Patterns of Finance for Higher Education: A Review of International Trends." Higher Education in Europe, Vol. 17, No. 1:141-149.
World Bank (1994) Higher Education: Lessons of Experience. Washington, DC: The World Bank.
Ziderman, Adrian and Douglas Albrecht (1995) Financing Universities in Developing Countries. Washington, DC: The Falmer Press.





12/29/00