Mingat and Tan (1985) have made attempts to measure the resources which would be
released by reducing or eliminating subsidies at higher levels. Their study indicates that
the scope for this tends to be greater in Francophone Sub-Saharan African countries
where unit cost differences are greatest. Using data from 10 countries (8 of which are
Francophone) they argue that 10% reductions in higher education subsidies would
permit about 2% increase in primary enrolments. Similar changes at secondary level
would support an increase of 1.6% in primary enrolments at the current unit cost levels.
It is only when large reductions in subsidy and high levels of recovery of operating
costs are introduced that substantial enrolment gains become possible at primary. Thus
if higher education student subsidies were cut completely, Gross Enrolment Rates
[GER] at primary could be improved by up to 18 points in some cases, though the
average is much less than this. Removal of all subsidies at higher and secondary level
and 100% cost recovery in higher education fails on its own to release enough resources
for universalising primary in most of the countries studied. Part of the reason is that the
countries with the smallest current primary enrolment levels also have the smallest
absolute enrolments in higher education and therefore the lowest amounts of subsidy
which can be transferred. Since unit cost ratios are much lower outside Africa the
impact of subsidy withdrawal elsewhere is potentially less. This situation would change
if higher education continued to experience the highest growth rates and pre-empt more
and more of the public budget.
Despite Mingat and Tan's analysis it should be remembered that worthwhile transfers
are possible. In a system which spends 20% of its finance on 1% of the cohort at
tertiary level a reduction in the length of higher education by one year from say four to
three (and/or by combining some income earning work experience with study to
achieve the same effect), or by increasing student teacher ratios by 25% (these were
estimated to be around 11:1 in 1983 in low income Sub Saharan Africa) could release
about 4% of the total education budget. This could represent as much as a 10% increase
in primary expenditures and enrolments. There are some examples where transfers of
this kind are being pursued (e.g. in Senegal and Ghana (Colclough with Lewin 1993).
A further dimension of the question of balance between investments at different levels
relates to secondary education costs. As primary GERs increase the cost burden of
secondary provision will increase if transition ratios from primary to secondary are kept
constant. Historically, transition ratios in most systems have grown as primary has
expanded thus tending to increase the proportion of total expenditure allocated to
secondary education. Secondary costs may grow at a multiple of increases experienced
during primary expansion since unit costs are typically about four times greater at
secondary. It then becomes at least as urgent a financial priority to examine efficiency
related reforms at secondary level that can reduce unit costs if heavy skews in favour of
secondary are not to emerge.
The implications of this brief analysis are that in some developing countries a case can