The Economics and Politics of
Income Contingent Repayment Plans *

D. Bruce Johnstone

To print this document, please use the PDF file.

Few ideas in higher educational finance have so caught the imagination of economists and politicians as the concept of income contingent loans for students. The essence of the income contingent loan is a debt instrument, the repayment obligation for which is expressed as a certain percentage of the borrower’s future income, rather than as a schedule of fixed payments over a defined period of time. All income contingent loans schemes, by definition, conform the burden of repayments to the borrower’s current income or earnings, and thus to his or her presumed ability to repay. All schemes must also assume that some number of low-earning borrowers will never be able to repay the true cost of their loans--not for defaulting, as such, but merely for an income that is never high enough or sustained enough to generate sufficient repayments to cover fully the lender’s cost of capital plus all other costs associated with the lending. As such, the income contingent loan tilts borrower subsidies not to all who borrow (as in across-the-board subsidized interest rates), nor to those borrowers whose parents were needy at the time the student had to borrow (as in means-tested low interest loans), but only (or mainly) to those for whom the personal investment in higher education, for whatever reason, has failed to “pay off” monetarily.

But at the same time, few concepts in the complex and high stakes arena of higher education finance have also been so poorly understood – or, expressed a bit more charitably, been held with such divergent understandings. As of the year 2001, several countries have apparently stable and successful income contingent loan programs: Sweden, Australia, and New Zealand come most quickly to mind, with the UK in the implementation process, and the US with an income contingent repayment mode as a relatively minor option. Many countries discussing the inauguration of large-scale student loan programs at the beginning of the 21^st century are evidently attracted to some kind of income contingent format. There is no question but that politicians, economists, and higher educational policy analysts in all countries and for the last thirty years or so have found the concept of income contingent student loans attractive – in fact, far more so than the relatively minor income contingent loan activity worldwide would suggest. The purpose of this paper is to explore three questions:

1.     What lies behind the enormous political and theoretical popularity of the income contingent repayment concept?

2.     To what feature of the entire student loan package (i.e. “income contingentness” or some other features) can one attribute the apparent success of the income contingent loan scheme in, for example, Australia (where its success is most widely touted)?

3.     What are the difficulties or other reasons for the income contingent playing so minor a role in many countries (particularly in the US) where student lending is large and generally available?

To begin at the beginning, it is necessary to remember that an income contingent loan is, first and foremost, a loan: that is, the borrowing of a claim on goods and services from lenders, or savers, to borrowers--for a price, expressed as a rate of interest. It is not absolutely essential to the concept of a loan that the money (or the claims lent by the savers to borrowers--in this case, to students) has to pass through the borrower’s hands. If a student is obligated to pay an amount of tuition, but is allowed (or even required) to defer that payment in return for a promise to pay a surtax on his/her income until a certain amount has been repaid at a certain rate of interest, then that student has borrowed and incurred an income contingent repayment obligation. The lender is most likely the government, which will have advanced the university (either from tax funds or from its own borrowing) the money that would otherwise have come to the university directly via the payment of tuition. In exchange the government receives an asset, much like a conventional promissary note, but in this case simply a promise to pay an income surtax. As in any lending, the present value of the anticipated repayments (in this case, of the anticipated surtaxes) must equal the amount lent (or tuition payments foregone) less any subsidy that the government is willing to provide.

If some borrowers, by reason of their low lifetime earnings, are not going to repay at the required rate of interest required to get lenders to part with their claims (i.e. their money), there must be subsidies, either from other borrowers, or from an outside source, presumably the government or taxpayers. This fact gives rise to the two general forms of income contingent plans, at least in the literature: the so-called mutualized plans, or those that recover some premium (measured by an excess percentage point or two or three of interest) from those who are able to pay, and the externally-subsidized plans, or those that cover the shortfalls on the low-earning borrowers from the government or taxpayer – just like other forms of student assistance.

The income contingent loan requires some kind of upper limit on repayments, such as an upper limit on the total effective rate of interest, an upper limit on the number of years in repayment, or an upper limit on the age of the borrower – or some combination of the above limits. Thus, the repayment obligation on a $10,000 loan, to be repaid on an income contingent basis at the rate of, say 5 percent of the borrower’s current annual income, could continue until the original loan is paid off at the effective rate of interest that the overall loan plan hopes to recover (say, 7 or 8 percent), or until 20 (or some other number of) repayment years have passed, whichever comes first.

Most borrowers will repay their loans at the target interest rate, with the repayment periods ranging up to 20 years. Those whose lifetime incomes prove to be permanently low will likely pay for the entire 20-year period and still have remaining obligations.  These remaining debts would thus be forgiven – sort of like grants based not on the low income of a student’s parents at the time of schooling, but on a student borrower’s low income for his or her lifetime after schooling.

A variant of the above would be repayment of the loan at full interest, or repayment for 20 years, or until the borrower attains the age of, say, 55. Such an upper limit, as in the preceding example, would protect the lifetime low earner, but would also protect – indeed, encourage – the nontraditional student borrower, who needs student loans in mid-life, but who might be fearful of a continuing repayment burden that might last into normal retirement years.

Still another variation on the upper limit is to require payment until all of the loans of a cohort of borrowers – e.g., all those beginning repayments in a given year – have been repaid at the lender’s full cost of money.  Such a scheme makes sense only when the intention is for repayment shortfalls from low earners to be made up by surplus payments from high earners in the same cohort of borrowers.  The financial integrity of the loan plan, at least for a particular subset of borrowers, is assured, providing there are a sufficient number of high earners in the cohort to balance the low earners.

This was the basis of the Tuition Postponement Option at Yale University in the early 1970’s—the first operational income contingent loan plan (which was abandoned by the University after several years of operation). The concept has the disadvantage of putting borrowers at some risk, depending on the probable future income or earnings composition of their borrowing class – and especially on how many potential high earners “opt” out of the ICRP for fear of getting into a cohort with too many potential low earners.  This is why more recent schemes have chosen to cover the shortfalls of low-earning borrowers from general tax revenues, obviating the need to follow cohorts of borrowers and reducing the incentive for potential high earners to opt out of the scheme altogether.  Conceptually, the weakness in the cohort mutualization concept has been the difficulty of demonstrating why the shortfalls from low earners, which admittedly must be made up somehow, should have to be made up not just by high earners – which would be not unlike any other scheme of income redistribution – but by that particular class of high earners who also had to borrow to finance their higher education.

The principle of income contingency makes sense, if only to provide some guaranteed protection to the student borrower who has trouble making loan payments because of low earnings and who does not want to default and incur damage to his or her credit rating. It is a more arguable proposition, at least as public policy, to employ income contingent loans as a kind of indirect income supplement for non-remunerative, but supposedly socially-desirable, careers. (Perhaps if these careers are so socially desirable, they should simply be paid enough to attract sufficient entrants.)

But it is when the inherent financial and legal complexities of income contingency are joined with the Byzantine political agendas of higher education that the real confusion begins.  The ICRP bed is full of strange fellows indeed. Some are economists. Many economists like income contingent lending because it tends to be associated not so much with a particular form of debt instrument, but generally (at least in the minds of these economists) simply with more student debt. And more student debt is good because it allows for higher tuition, and for the higher education cost burden thus to be shifted from the taxpayer to the student.  And that is thought to be a good thing because the student-university relationship then more nearly resembles a market transaction.  And that is good because that is the way God intended things to be. Or would have if He had fully grasped the principles of economics.

More specifically, the further shift of costs from taxpayer to student – again, presumably abetted by the existence not just of student loans, but by the larger debts thought to be sustainable by loans of the income contingent variety – is thought by the economist to be more efficient because it is thought to make the seller (the university) more responsive to the buyer (the student), both in the kind and the quality of the education provided.  It is further thought to be more equitable because the average student is known to enjoy both more affluent parents and more promising income prospects than the average taxpayer.

Most of the economists so enamored with ICRP are thus enamored less with the peculiarities of the income contingent loan contract itself, and more by a possible associated shift in the cost burden from the taxpayer to the student and by the consequent enhanced role of market behavior in guiding the fortunes of higher education.  But there is another economist fellow in the ICRP bed, often more theoretically inclined and more removed from the details of implementation, who is enamored precisely with the form of the ICRP instrument itself.  He or she admires its resemblance to equity financing: the raising of human investment capital by selling shares, as it were, in the borrower’s life income prospects.  ICRP becomes almost a metaphor of a proper market.

Then there are in that bed the proponents of pricey private colleges and universities, mainly but not exclusively in the US.  These institutions need students to be able to handle very large debt loads, and they think that ICRP loans will make the big debt burdens less burdensome.  But even more, they like ICRP because they, like the economist in the foregoing paragraph, associate ICRP with a shift of cost burden from the taxpayer to the student, ushering in the dream of all private colleges and universities, which is full -- or near full -- cost tuition in the public sector, combined by generous and portable need-based grants for the poor, thus erasing the price advantage that public institutions currently have over their private brethren and sending the most desirable students back to the private sector, where they are known to belong.  Little short of the privatization of the public sector.  Or perhaps the publicization of the private sector.

There are also some citizens and politicians, generally of a conservative stripe, in that ICRP bed, seeing it as one more weapon in the arsenal of lower taxes and less government.  They care less about efficiency or equity, but their prescriptions are still for high tuition, and they, too, see income contingency as abetting their cause.

But then there are also lots of financial aid folks and others in the ICRP bed who don’t care a fig for the privatization of higher education, but who simply see more and more students struggling with high student debt loads and uneven income prospects, and who believe that the conformance of repayment burden to lifetime income or earnings is a sensible idea, particularly when combined with a provision to provide real relief for the student borrower who repays at a maximum percent of income for the better part of his or her working life, and who arguably is every bit as entitled at some point to a need-based grant as is the student whose parents merely happen to be poor while the student is in school.  I hope that many of most in this audience fit this model.

There is nothing wrong in making public policy with a mixture of motives – or, to stick with the nautical metaphor, in a crowded and not always harmonious boat.  But those who advocate income contingent student lending should be clear, at least to themselves, what it is they are advocating.  Most importantly, there are a number of theoretically legitimate, if highly controversial, reasons for advocating a major shift in the cost burden of higher education from the taxpayer to the student.  I personally do not buy into those reasons, and at another time and place would be happy to explain why to anyone who should care to listen.  But such a truly momentous shift in public policy – from one, in both our countries, of sharing the costs among students, parents, and taxpayers, to one of placing the overwhelming burden upon the student, or at least upon the student who is without parents able and willing to assume it on his or her behalf – is far, far too important a debate to allow to be obscured by the myths and misperceptions that have traditionally surrounded the ICRP debate.

What are these myths and misperceptions?  I will suggest five, reminding the audience that I began this address with the stipulation that I believe the ICRP concept to be a good and quite implementable idea.

1.     An ICRP debt is still a debt, which the borrower, or at least most borrowers, must repay at the same real cost – measured in simple annual interest rates or, in economics talk, at the same present value – as any other student debt.  Unless, of course, the government chooses to subsidize ICRP more than it subsidizes alternative, conventional student loans.  But then any loan program that is the object of more government subsidies, or taxpayer largesse would presumably be made more attractive to the student borrower.  For the overwhelming number of student borrowers, then, the income contingent loan may be more convenient – but it will not be cheaper.

2.     Any special convenience, whether to the borrower or to the lender, that is attributed to the assumption of collections by the general income tax withholding system or to any other pervasive governmental collection mechanism, such as the US Social Security system, should not be credited to the concept of income contingency.  If the government wants to collect student loan repayments through income tax withholding, it can do so – for any kind of student loan.  Such a provision, I believe, makes sense, although this is also a door that our respective governments should open only with the greatest care lest it seem too attractive to too many good causes looking for a more effective way to ensure their allegedly worthwhile payments or collections.  Again, this is not an argument against ICRP: only a caution against imbuing income contingency with virtues that are not truly its own.

3.     Don’t assume that ICRP will eliminate all defaults.  To the degree to which defaults are a function of unduly burdensome repayments, an ICRP will at least lessen the basic cause, although special forbearance provisions for those in repayment status and out of work or otherwise earning very low incomes can be incorporated in any student loan program.  But to the degree that defaults are a function of a bad college experience, or to frustration over the inability to find a job, or to any number of other personal problems that beset young people today, or merely to the incredible rootlessness of many youth, defaults will continue--ICRP or not. Once again, this is not an argument against ICRP, nor a denial that defaults, as we conventionally measure them, will probably go down in any well-administered ICRP.  It is, rather, just another plea for us to place the presumed virtues of ICRP in a little better perspective than so frequently advanced by its true believers.

4.     Don’t think that ICRP is administratively simple just because it is superficially simple in concept.  All you pay is a percent of earnings, right?  Sure.  But what earnings?  Last year’s taxable income?  This year’s estimated taxable income?  All taxable income, including such elements as capital gains and other forms of attributable income, or just earnings – which is, after all, that portion of income most clearly attributable to education?  Is it to be individual or joint income, and what do we do about what could become yet another deterrent to the once important institution of marriage?  And what of the borrower who again becomes a dependent of his or her parents?  It is not that these and other similar questions are so difficult, or that, once answered are so very hard to implement: it is just that implementation is so very much more complicated, and so very much more dependent upon non self-evident answers, that it often seems upon first hearing the concept.

5.     Finally, do not assume that the students whom you or I may assume to have the most uncertain future income prospects, and thus whom we might assume to be the most eager and accepting ICRP borrowers will be so at all.  Like it or not, both of our countries (indeed, all of the world) continue to exhibit very significant intergenerational transmission of socioeconomic status and earnings power.  Without the jargon, this simply means that the children of the well-off are likely to be so too, and the children of the poor, and especially the very poor, are quite likely to be poor also.  And children of the poor, for all sorts of reasons, do not particularly like debt, even though they may have to incur a good deal of it.  Once again, this is not an argument against ICRP.  But do not look to students to provide the political fuel for an ICRP policy, and especially do not look to low income children or low income families, who are as likely to perceive ICRP as a slick government trick to get out of its rightful obligation to provide them a low cost public higher education.

What, then, is the verdict on ICRP – economically and politically?  The first thing to remember is that the economics and politics of ICRP are hopelessly entangled.  The next thing to remember is that ICRP loans are much more like other student loans than they are unlike them – and that the student borrower must repay them, like other loans, and that the taxpayer will almost assuredly have to subsidize them, to some degree, like other forms of student assistance, including conventional loans.  ICRP loans are not simple, either to understand or to implement, and it probably makes little sense to begin an actual program until student debt loans have become both substantial and extensive.  Finally, the public policy of allocating the burden of higher education’s costs among parents, students, and taxpayers – and particularly if there is to be a massive shift of burden from taxpayers to students – should be faced squarely, not slipped in under the guise of alleged student loan reform.

Then, in its proper perspective, Ontario and desirably all of Canada, and certainly the United States, should implement a provision for students, if they choose, to repay their student loan debts at an annual rate not to exceed more than a certain maximum percentage of their incomes.  It will be a significant, if not an earth shattering, reform.

Income contingent Loans: A Selected Bibliography

Albrecht, Douglas & Adrian Ziderman (1991) Deferred Cost Recovery for Higher Education: Student Loan Programs in Developing Countries. World Bank Discussion Paper No. 137. Washington, DC: The World Bank.

Barr, Nicholas (1989) Student Loans: The Next Steps. Edinburgh: Aberdeen University Press.

Brody, Evelyn (1994) “Paying Back Your Country through Income-Contingent Loans.” San Diego Law Review, 31:2 Spring 1994, pp. 449-518.

Johnstone, D. Bruce (1972) New Patterns for College Lending: Income Contingent Loans.  New York and London: Columbia University Press.

Johnstone, D. Bruce (1972) “The Role of Income-Contingent Loans in Financing Higher Education,” Educational Record, Vol. 53, Spring 1972, pp. 161-168.

Johnstone, D. Bruce, Daniel B. Wackman, and Scott Ward (1972 ) “Student Attitudes Toward Income Contingent Loans,” The Journal of Student Financial Aid, Vol. 2, No. 1, March 1972, pp. 11-27.

Johnstone, D. Bruce and Preeti Shroff-Mehta (2000) “Higher Education Finance and Accessibility: An International Comparative Examination of Tuition and Finance Assistance Policies.” Inter-national Comparative Higher Education Finance and Accessibility Project, Center for Comparative and Global Studies in Education, State University of New York at Buffalo. http://www.gse.buffalo.edu/org/IntHigherEdFinance/

Swedish National Board of Student Aid (CSN). (1997) A Report on Some Effects of the Present System of Student Loans (October 23, 1997). [Stockholm]: Staff of Director General.

Woodhall, Maureen (1989) Financial Support for Students: Grants, Loans, or Graduate Tax? London: Kogan Page (in association with the Institute of Education, University of London).

Ziderman, Adrian and Dougles Albrech (1991) Deferred Cost Recovery for Higher Education: Student Loan Programs in Developing Countries.  World Bank Discussion Papers. The World Bank. Washington, D.C.

Ziderman, Adrian and Douglas Albrecht (1995) Financing Universities in Developing Countries.  The Stanford Series on Education & Public Policy: 16. The Falmer Press.  Washington, DC.

Revised 3/12/01

To print, please use the PDF file.

---