International student funding comparisons: conclusions

The strategic conclusions from country experience can be summarized as follows. The USA has a useful model on tuition fees, but still lacks a good loan scheme. The UK, after a bad start with mortgage-type loans, now has an effective loan scheme, but major reform is still needed to reduce complexity, to address the heavy costs of interest subsidies, and to phase in flexible fees.

The Netherlands and Sweden, rather like the UK, have loans that (implicitly or explicitly) are in-come contingent, but both have flat-rate tuition fees fixed by government (in Sweden fixed at zero). Australia has a well-established loan scheme, but has not avoided interest subsidies, and has got into a muddle on fees. New Zealand offers useful lessons on loans and fees and, more recently, an example to avoid on interest subsidies.

Helpful features
Earlier discussion suggests a number of useful strategic design features. Where possible, incentives have advantages over regulation in encouraging efficient behaviour by all actors, students, universities, employers, and government. Enforcing regulation can be costly and administratively demanding. Incentives are likely to be cheaper and more effective. Secondly, it is desirable to avoid price subsidies, which are generally costly and regressive. One implication is a move towards market-determined fees; another is a move towards unsubsidized interest rates for student loans (though any such moves have a major political dimension). Thirdly, income testing should be avoided where possible: it creates a major administrative burden and risks creating ad-verse incentives.

A fourth lesson suggests developing flexible systems that can evolve. It is highly undesirable to bring in a new system and then have to change it drastically (as with the UK loan system). Any change of system is costly and disruptive. In the face of uncertainty, a rational strategy is to design a system that encourages diversity and is capable of evolving, pointing towards a strategy of market forces plus regulation, rather than one of central planning, which is better suited to a predictable and more static world. The system should therefore include loan schemes that are able to grow, to have their parameters (for example, the interest rate charged to borrowers) changed, and to be extended to further groups of students. Tuition fees should be able to vary across subject and across institution. The amount of public subsidy for higher education should also be capable of varying. University governance should make it possible for course content and degree structures to evolve as circumstances change.

Pitfalls
Economic theory and practical experience point to avoidable problems: (a) fiscally unsustainable public spending; (b) public spending hijacked by the middle class; (c) loans absent, or badly designed, so that they bring in few, if any, extra resources; (d) economic constraints on education providers, which re-duce incentives to efficiency; and (e) specific design features that are costly (interest subsidies), administratively demanding (income testing), or both. These occur in all the countries listed above, though (b) and (d) are less of a problem in the USA and New Zealand, which have variable fees. They also occur elsewhere: an account of Latin America reported that: Most of the public institutions . . . have argued that low or no tuition fees have provided greater equality of educational opportunity by providing greater access.... Such reasoning is simply incorrect . . . the overwhelming public subsidy has been and continues to accrue to students from middle and high-income families.

Implementation
Alongside policy design, a final, and critical, lesson is the importance of implementation, in many ways the harder part of the task. Discussion here concentrates on student loans, but applies generally.

The first question for reformers is whether key institutional prerequisites are in place, including the capacity to collect repayments and a legal system robust enough to enforce them (both points apply equally to mortgage loans and income-contingent loans). If either is significantly inadequate, the idea should be postponed. If reform gets the green light, it is essential that implementation is taken fully into account at the time the policy is formulated. To that end, the group of people who put together any reform package needs to contain both policy-design and implementation skills. The need for strategic thinking about policy requires little elaboration, since that is what this entire book is about. In the case of loans, it is needed, for example, to ensure that interest subsidies and income testing are not introduced unthinkingly. Political skills are a second essential. They are needed to communicate why a loan scheme is necessary and how it can help access, or to explain the role and operation of tuition fees.

UK experience over the 1990s shows how failure to grasp this point brings students onto the streets. It is also necessary to ensure that political support is continuing. New Zealand is an example of how a reform can be introduced successfully but subsequently at least partly undone. A wide-ranging set of additional skills address the operational needs of the scheme: the obvious starting point is expertise in running a system with large numbers of loan accounts that remain active over many years; for income-contingent loans, tax expertise is needed to ensure that the scheme is compatible with the operations of the tax authorities; if loans are to be paid using private money, deep financial market expertise is necessary, and so is an understanding of the classification problem; if loans are to be paid to students through banks? networks of cash points, expertise in retail banking is needed.

The UK experience shows that attempts to involve banks at a late stage in policy design without any awareness of how they operate can poison relations with private providers for a considerable period. Finally, there has to be sufficient time for any new system to be implemented effectively.