The research addresses the question:

Using Scotland and Wales as comparators, how do different approaches to universalism in undergraduate funding affect the distribution of student loan debt and what implications might this be expected to have for the reproduction of economic inequality within the graduate population?

Following Esping-Andersen’s (1990) development of a typology of welfare state forms, a body of literature has developed using empirical data to examine the effectiveness of targeting benefits by income as a means of addressing inequality.  Korpi and Palme (1998) argued that under the “paradox of redistribution”,  systems which concentrated benefits more on those from lower incomes  were less effective at addressing inequality than those characterised by a greater level of universalism.  These findings are controversial (Gugushvili and Hirsch 2014): they have been contested by some studies (for example, Kenworthy 2011, Marx et al 2013) and supported by others (most recently, Jacques and Noel 2018).

This debate tends to omit student funding, with rare exceptions (Sefton 2002). However, as an increased number of people enter higher education across Europe (OECD 2018) student funding is becoming a critical part of welfare support to young people as they transition to adulthood (Antonucci 2016), as well as those entering higher education only in adulthood.

In order to support increased student numbers, governments in many jurisdictions have begun to use to student loans (Johnstone and Marcucci 2010), including within Europe (Antonucci 2016). These introduce a further complexity into debate about the relationship between welfare benefits and inequality, as loans combine short-term benefit with potential future costs (Hunter Blackburn 2018). How these longer-term costs are distributed by income will be influenced by what choices governments make about how they share repayable loans and non-repayable funding, such as tuition fee subsidies and living cost grants, among students from different incomes. It may also be influenced by students’ later earnings, where “income-contingent” repayment schemes are used (Barr 1991), as in the UK.

My research examines how different choices about the design of loan-using student funding systems affect the distribution of long-term costs among participants, according to their initial family income. It takes advantage of the devolution of student funding within the UK since 1999 to compare the effects of divergent choices in Scotland and Wales about the balance between universalism and targeting for non-repayable student funding by income. Decisions on that have in turn affected how far each government has to rely on student loans to provide higher levels of student funding to those from lower incomes. 

In Scotland, the priority for non-repayable funding has been supporting a policy of universal free tuition, so that no fee charge is levied to any first-time, full-time undergraduate: living cost support has been provided almost entirely through loans. In Wales, a smaller proportion of non-repayable funding has been used for universal fee subsidies, and a larger amount directed towards living cost grants targeted on these from low incomes, reducing how much their living cost support is provided through loans.

After describing the different policy choices made in each nation and how these are reflected in the funding made available to students, the research question is explored by considering

-       the distribution of actual borrowing by initial family income in each nation by the end of study,

-       what effect the income-contingent repayment system used in both nations is expected to have on how far initial borrowing translates into actual repayments, according to later earnings, and

-       how the two systems therefore compare in the extent to which their student funding systems function as a redistributive form of welfare, once post-graduation repayments are taken into account.

The research is the first to undertake a quantitative analysis of administrative data from student support providers for Scotland (the Student Awards Agency Scotland) and for Wales (the Student Loans Company). Data for a complete cohort leavers from full-time, first-time higher education are used for each nation. The data are available for short- and long-cycle undergraduate study, which are generally considered separately in the analysis below. The distribution of total borrowing by income in each nation is first compared using measures of central tendency and distribution. The research then explores how far any differences between the two nations may be due to differences in students’ take-up of student loans, rather than the structure of income-related entitlements alone. Logistic regression is used to investigate the relative strength of the relationship between students’ borrowing behaviour and individual characteristics (income, level of study and other factors available from the data) in each nation. An estimate is then made for the distribution of expected total real value of repayments according to income, based on the figures calculated for final borrowing. An income-related loan repayment model, adapted from one developed by the UK government for use in England, is used to estimate the real value of repayments for different levels of borrowing, according to a person’s earnings decile, taking account of variations in the repayment rules in the two nations. The expected distribution of students across the earnings range is estimated, drawing on a variety of recent UK research into graduate earnings according to student characteristics (particularly Britton et al 2016). These two analyses are then used together to estimate the distribution of the real value of repayments in each nation, by initial household income. This is compared with the distribution of initial borrowing, to assess how far this remains relevant once the repayment rules are applied. As the final stage, the results are converted into a form which can contribute to the wider debate on targeting and redistribution by calculating the Gini coefficient for each system (a) for the expected real total value of loan repayments (b) for non-repayable benefits received during study and (c) for the net effect of these. This will enable the outcomes of the two systems to be compared directly for their long-term redistributional effect, taking income on entry as the measure of initial population inequality.

Initial results show that student loan debt is relatively evenly distributed by income for students from Wales, with borrowing skewed slightly towards those from higher family incomes. The variation in borrowing within income groups is limited. In Scotland, differences in borrowing between income groups are more marked, with those from the lowest incomes borrowing substantially more than those from higher incomes. Borrowing also varies more within income groups in Scotland, an effect which increases as income rises. Figures for Scotland are revealed to be more influenced by student borrowing behaviour than those for Wales. Those borrowing nothing at all are de minimis in Wales: but 15% of those in Scotland borrow nothing, and a larger proportion than in Wales take out only some of the loans available to then. In Scotland income and other measures of socio-economic advantage have a more substantial effect on take-up, as do the likelihood of living at home, level of study and other factors to a more limited extent. Examining the effect of the repayment systems shows that in both nations for lower earners differences in initial borrowing are largely irrelevant, with differences becoming apparent in the upper half or so of the graduate earnings distribution. As a result, repayments become closely dependent on initial borrowing at higher earnings. These results imply that, as it results in higher borrowing from those from lower incomes, Scotland’s more universal system is more likely to reinforce initial differences in income over the long term than that for Wales. However, the effect of this is contained among mid-to-higher earners. This should reduce the impact on the Gini coefficient for Scotland (still to be calculated) as higher earning graduates tend to be drawn from higher income families.

Antonucci, L. (2016). Student lives in crisis: Deepening inequality in times of austerity. Bristol: Policy Press. Barr, Nicholas (1991) Income-contingent student loans : an idea whose time has come, in Shaw, G K, (ed.) Economics, Culture and Education - Essays in Honour of Mark Blaug. Aldershot: Edward Elgar Publishing. Britton, J., Dearden, L., Shephard, N., and Vignoles, A. (2016). How English domiciled graduate earnings vary with gender, institution attended, subject and socio-economic background. London: Institute for Fiscal Studies. Esping-Andersen, G. (1990). The three worlds of welfare capitalism. Cambridge: Polity Press. Gugushvili, D. And Hirsch, D. (2014). Means-testing or universalism: what strategies best address poverty. A review contributing to Joseph Rowntree Foundation’s development of an anti-poverty strategy. Leicestershire: Loughborough University, Centre for Research in Social Policy. Hunter Blackburn, L. (2018), Student Support in Wales: A Case of Progressive Universalism?, in Sheila Riddell , Sarah Minty , Elisabet Weedon , Susan Whittaker (ed.) Higher Education Funding and Access in International Perspective (Great Debates in Higher Education, Volume ) Emerald Publishing Limited. Jacques, O. and Noël, A. (2018) The case for welfare state universalism, or the lasting relevance of the paradox of redistribution. Journal of European Social Policy, 28(1), 70–85. Johnstone, D. B. and Marcucci P. N. (2010). Financing higher education worldwide : who pays? who should pay? Baltimore: Johns Hopkins University Press. Kenworthy, L. (2011). Progress for the poor. New York: Oxford University Press. Korpi, W. and Palme, J. (1998). The paradox of redistribution and strategies of equality: Welfare state institutions, inequality, and poverty in the western countries. American Sociological Review 63(5) 661-687 Marx, I., Salanauskaite, L. and Verbist, G. (2013). The Paradox of Redistribution Revisited: And That It May Rest in Peace? Bonn: The Institute for the Study of Labor (IZA). OECD (2018). Education at a Glance 2018: OECD Indicators. OECD Publishing: Paris. Sefton, T. (2002). Recent changes in the distribution of the social wage. CASE paper 62 London: Centre for Analysis of Social Exclusion, London School of Economics.