Fairest of them all? Comparing the reproduction of inequality in student funding in Scotland and Wales

Abstract

This research is concerned with how ideas of fairness are applied to student funding systems. The background to the research is the adoption of student loans in many countries in recent years. The argument is made that in judging whether student funding systems which make use of loans are fair, more use should be made of the concept of the reproduction of economic inequality and that greater attention should therefore be paid to how such systems distribute their direct financial benefits and costs according to students’ initial family income. The research makes use of the devolution of student finance in the UK to compare the short-term and expected long-term distributional effects of divergent policy choices in Scotland and Wales in relation to the targeting by income of repayable and non-repayable types of funding. It draws on individual administrative records held for Scotland by the Student Awards Agency Scotland (N=28,503) and for Wales by the Student Loan Company (N=12,557), covering a complete recent cohort of full-time first-time undergraduate leavers in each nation.

Such records have not been used before in the United Kingdom for the purpose of examining the detailed operation of student funding systems. The research makes a new contribution to the literature on student funding by using these records to establish how such funding is taken up in practice. It makes innovative use of concentration curves and indices, commonly used in empirical welfare literature. These are used to compare the short-term and long-term distributional outcomes of student funding policies in the two nations studied. The findings show that while both systems skewed the distribution of total initial benefits, in the form of loans and grants combined, towards those from lower incomes, initial borrowing was skewed towards those from lower family incomes for Scotland and towards those from higher ones for Wales. While this was predictable from the design of each system, differences in the take-up of student loan by income are also shown to be relevant; loan take-up was lower among students from Scotland than those from Wales and inversely related to income.

Average marginal effects estimated from logistic regression models are compared to show that differences between the two nations in the relationship between borrowing behaviour and income persisted after taking differences between the two student populations into account. Turning to costs, in the form of loan repayments, a range of earnings scenarios are developed for the populations studied to examine the expected distribution of these. For Wales these are shown to be highly predictably skewed towards those originally from higher income families. The distribution of costs relative to initial family income is more unpredictable for Scotland, and depends on assumptions made about the level of the loan repayment threshold and the operation of the labour market. For young degree-level students, a recent increase in the threshold is predicted to make it less likely than before that repayments will reproduce the skew of initial borrowing towards those from lower incomes. Costs are also predicted to be less likely to skew towards this group, the more closely higher initial family income predicts higher later earnings. “Nil income”, largely independent (mature), Scottish students are however expected to face higher average repayments relative to others under all assumptions examined; repayments are also predicted in all circumstances to skew towards those from lower family incomes within the Scottish sub-degree graduate population. The results suggest that the political claims made for both systems tend to concentrate on government inputs rather than outcomes for graduates and that expected outcomes are more dislocated from political claims in Scotland. The research raises normative questions about what redistributive effects student funding systems should be expected to have and the relevance of students’ initial family circumstances to their later repayments, and also offers insights into the wider unused potential of national administrative student funding data sets as a research resource. The results further provide the basis for theorising more generally about the conditions under which student loans systems might be expected to increase or decrease the reproduction of economic inequality within the graduate population over the life course, and under what conditions predictions of long-term effects can be most securely made.