European Universities Must Get Their Act Together
Published on: Sep 12, 2006

Universities should rely more on private funding 

Gap between social and private returns is small and declining
Each additional year of education, typically, raises wage incomes with 5-10 per cent. These returns are generally larger for higher education. If social returns exceed private returns, education causes positive external effects to society and the government should support education. Estimating macro-economic production functions where total output is explained by human as well as physical capital, one obtains macro returns to education of about 5-6 per cent for each year of education. This is at the lower end of the estimated micro returns. Despite widespread belief in large externalities of education, social returns seem slightly lower than private returns.
However, empirical findings suggest that private returns to higher education are substantial. A popular argument is that the government should expand investment on education rather than reduce public debt, because the private returns from study are higher than the safe real return on government bonds. But the government should intervene in higher education because the social exceeds the private return to education not because private returns are large. The returns on education are higher than on government bonds because human capital is illiquid and more risky as labour incomes fluctuate due to business cycles, sectoral shifts, technological developments, international trade, etc. If skilled graduates earn higher incomes than low-skilled workers, it is profitable to invest in higher education. Especially, the US and the UK have experienced dramatic increases of the skill premium.

Baumol’s cost disease also suggests more private funding

    Teaching and research need to be done by highly qualified people and cannot be replaced by technology. Productivity growth in universities inevitably lags behind, so the cost and price of a university education rise over time. This does not warrant a growing subsidy, since the increase in productivity elsewhere boosts purchasing power. Skill-biased technical change boosts the returns to study. Also, if higher education is a luxury good, it flourishes as technical progress makes people wealthier. Graduates can thus rationally use the higher returns to pay for the higher cost. Provided the opportunity costs of study do not increase as much as tuition fees, Baumol’s cost disease expands the university sector. Hence, despite rising relative prices, the budget share of higher education rises over time.

To conclude, the crisis of European universities is not due to lack of public funds. There is no evidence that the social return to study exceeds the private return sufficiently to warrant bigger state subsidies. If anything, the private return to higher education seems to be rising as may be witnessed from the growing skill premium that graduates command in the market. However, higher education in many parts of Europe is starved of funds. The lack of funds will worsen due to the relentless operation of Baumol’s cost disease. Much more can be asked from students provided they can make use of income-contingent loans. Even though student poverty is a real issue, graduates are relatively well off.

Misguided equity motives in higher education 

Empirical research suggests that the ability of the student and long-run background factors (‘culture’, ‘family’, ‘environment’) are the most important determinants of enrolment in higher education. Increasing enrolment in higher education of children from lower socio-economic backgrounds therefore requires intervention in basic and secondary education rather than generic subsidies for higher education. Equity grounds for large-scale subsidies to universities are doubtful. The vast majority of students in higher education belong to the richest half of the population. Moreover, the average tax payer has less lifetime income than the average graduate. All kinds of politicians raise equity issues for the wrong reasons.
Some argue that university education is a ‘basic right’ and should be free of charge. Universities should be accessible to all with sufficient academic capabilities. But this does not imply that higher education should be free from charge, neither does it imply that all should pay the same price, or should pursue the same quality of education. Another misguided argument is that subsidies are good as graduates pay more taxes. But the extra tax revenues do not recoup subsidies as most governments over-subsidise education (De La Fuente and Jimeno, 2005). Also, high-income earners who do not study do not receive subsidies, but still pay higher net taxes compared to those who do study. The poor may benefit from regressive higher education subsidies as they allow the government to use the progressive income tax at lower efficiency costs. Education subsidies reduce the tax distortions on human capital investments. The costs of study should therefore be tax deductible, but not the interest as this induces over-investment and distorts saving.
Some politicians reject ‘elitist’ universities where the brightest students receive the best and most expensive education. This boils down to a plea for high taxes on investments in higher education and thus obstructs profitable investments in human capital. The best students migrate abroad and individuals with lowest incomes are worse off than with direct redistribution. Both efficiency and equity are harmed by holding back talented students. Low tuition fees should not be used for equity reasons either, since it is inefficient to tax study at 100 per cent above the fixed tuition fee for those who want to pay. Income redistribution should be carried out through the tax system and not through the education system. Too low tuition fees erode the tax base by causing under-investment and the poor are eventually worse off than with more progressive taxes. If the purpose of low fees is to guarantee access to universities, and not income equality, an income-contingent loan scheme is sufficient.

From student grants towards income-contingent loans 

Capital markets fail to deliver loans to finance tuition and costs of living as banks cannot easily assess the risks of some students and face difficulties monitoring efforts by students and graduates. The resulting adverse selection and moral hazard effects give rise to high interest rates, credit rationing or even a collapse of the credit market for student loans. In addition, students are risk averse and hesitate to take up large loans. Indeed, risks associated with study cannot be insured due to incomplete contracts and information problems. Imperfect capital and insurance markets induce underinvestment in higher education and hurt especially more loan-averse students from poorer backgrounds. Such students are forced to work, disturb the quality of teaching and more frequently dropout. Hence, there is a case to help such students so that they can pay higher tuition fees.

Income-contingent loans rather than student grants and subsidised tuition

To tackle student poverty, students should be allowed to borrow for fees and cost of living. Income-contingent loans (ICL) can overcome problems of capital market imperfections with risk-averse students. ICL only require students to pay back principal and interest if their incomes after graduation are high enough. ICL thus offer a combination of loans and social insurance. If income risks of graduates are pooled, fewer subsidies are needed to eliminate risk aversion.
Commercial banks and insurers are unable to write contracts based on future incomes, but the government can enforce contracts through the tax authorities and verify earned incomes. By selection and tracking of student performance and denying funds to non-performing students, the government can more easily eliminate the ‘rotten apples’. It can also collaborate with other tax authorities in Europe to track down graduates who try to default. In principle ICL feature no subsidies. However, the risks of default may be borne by society. ICL avoid perverse redistribution from the average taxpayer to students, because the majority of students comes from higher income classes and will belong to the higher income classes after graduation.
An alternative is a graduate tax (GT) where graduates receive grants financed by issuance of government debt. Graduates repay a fraction of their lifetime incomes. The government pools this income to repay government debt including interest. From the individual perspective, repayments under a GT can exceed loans (including interest) as graduates with high incomes under a GT typically pay more. A GT thus has more insurance and redistribution than ICL. In practice, there is only a gradual difference between a GT and ICL. Under a GT repayments by high-earning graduates exceed the costs of their education and the surplus is used to subsidise low-earning graduates. If a GT is budgetary neutral, it is like ICL with risk pooling. In the absence of moral hazard, a GT provides more insurance than ICL and thus dominates a pure loan. With moral hazard, however, ICL provide better incentives as it features less insurance and performs better than a GT if risks are pooled among students and not borne by the government.
Both ICL and a GT distort labour supply and encourage delay of career choices in order to avoid repayments that are contingent on future incomes. Students may not put enough effort in studying hard; they may study longer or enrol in ‘fun’ studies. These moral hazard problems can be avoided by selection and penalties for those who do not make satisfactory progress. A bigger loan warrants a higher tariff. This prevents cross-subsidies from cheap to expensive courses and avoids income redistribution from smart (high return, low risk) to less bright (low-return, high-risk) students. As a result, there is less moral hazard and more pure insurance. To prevent cross-subsidies from profitable to loss-making studies, tariffs per course and per discipline must be differentiated (see section 4). We prefer ICL to a GT, because they feature less insurance, allow more flexibility in repayment, and can be better tailored to avoid moral hazard. This is especially the case if repayment parameters are not very differentiated by size of loans, type of study or student performance. In that case, the GT causes a potentially large moral hazard problem as the link between funds received and repayments is weakened a lot.
Insurance of default risks may also give adverse selection. Rich students may avoid ICL or a GT to avoid risk pooling, except if the government finances the cost of bad debtors out of general funds rather than a surcharge on interest. These transfers benefit only students with very low lifetime incomes. An alternative is to make participation in ICL or a GT obligatory. Adverse selection also arises if talented but ‘poor’ youngsters do not participate due to loan aversion and work rather than study. Good information may convince them that it pays to study and that they do not run large income risks if they finance their studies with ICL.