The combination of spending cuts, efforts to protect the poorest from some of their effects, and ‘localised’ decision-making are leading to an increase in the numbers of means tests designed by lower level institutions. This paper examines a case study of the effects of this, looking at the means-tested support which has been offered to English students applying to go to 52 universities from Autumn 2012, designed partly to offset the rise in general fees to or towards £9,000. 27 of these universities are offering significant levels of means-tested support through bursaries or fee reductions depending on parental income. Although using a common income definition, each university has designed its own system with widely varying criteria. Taken with the national maintenance grant system, these imply substantially different levels of support for students from lower and higher-income families. Nearly all of them involve significant downward steps or ‘cliff edges’ in support at particular income levels, often involving a drop of several thousand pounds. Just looking at student support by itself, this implies typical marginal withdrawal rates exceeding 40 per cent and even 100 per cent over particular income ranges. Taken together with the effects of other parts of the tax and benefit system that would have affected family income, the effect is marginal effective tax rates on family resources that exceed 50 per cent for the entire income range up to £43,000. They typically reach 180 per cent for £1,000 income variations around income of £25,000. For the most generous case, Oxford University, the £13,050 reduction in subsequent means-tested student support is equivalent to nearly the whole of the £13,250 net income difference between two-child families that earned £17,000 and £44,000 in 2010-11. As well as introducing extra complexity and questions of equity in treatment within student finance, this kind of development runs counter to other parts of government policy, such as Universal Credit, intended to smooth out and simplify means-tests. As localisation is pushed further, and more agencies become responsible for designing their own mean tests, the lack of a system to take an overview of their overlapping effects, and to avoid undesirable design features will become an increasing problem across social policy, not just in this particular area.
