Intergovernmental grants that aim to remove all inequality between communities protect bad governments from the consequences of their actions, and interfere with the useful demonstration of the relative strengths and weaknesses of policies and programmes which result from local control.
The second alternative is far preferable to the first. Studies in the USA show that when poor areas receive transfers over which they have control, they make better spending decisions than when a big jurisdiction dictates the use of subsidies to them. The fact that they are unable to raise sufficient revenues locally should not be used as an excuse to deprive poorer areas of the political right to determine their own needs. They should receive grants according to a formula, with no provisos, so that they can spend the money according to local priorities.
Some governments may wish to introduce a tax credit system whereby any citizen who contributes to a neighbourhood non-profit development trust or corporation receives a credit against national, regional or local taxes of, for example, between 10% for high-income earners and 80% for low-income earners. This would mean that development programmes compete for citizen support rather than government support.
The federal government works out a Representative Tax System (RTS) based on a weighted average of provincial taxes including 39 separate types of taxes, rents and user fees. This is arrived at by dividing the total revenues from each tax by the total tax base. For example, the total property tax receipts divided by the total property tax base of the country comes to 1.65%. To determine the RTS the same calculation (revenues divided by tax base) is applied to every tax on an annual basis.
The RTS is a fictional system and none of the provinces use it, but if it were applied to the tax bases of all ten provinces it would yield the same total as that which they collect from their own diverse systems.
The government's accountants average out the amount the five middle provinces would raise per capita if they charged taxes at RTS rates. They then apply the RTS to each province's taxable resources and make up the difference between the potential RTS revenues of an average province and the potential RTS revenues of any province that falls short of the average, with a single, no-strings-attached, annual grant.
Federal equalisation grants are not related to the way in which the provinces raise their own taxes, and those that receive grants are not told how to spend them. But the system does ensure that each province could, if it wanted, raise sufficient resources to meet public needs at a level which is average for the entire country by applying the RTS and adding on the grant. Even though four of the ten provinces (Ontario, Alberta, Saskatchewan and British Columbia) do not receive any RTS payments, the system is well accepted in Canada.
This kind of system can be applied between first- and second-tier, and between second- and third-tier, levels of government. In other words, national governments can provide equalisation grants to poorer regions based on RTS assessments, whilst regions use a similar system for redistribution among local communities. At the regional level, equalisation grants should be paid annually, but within regions, subsidies might be paid bi-annually or quarterly.
If such a system were introduced in South Africa, the Central Witwatersrand metropolitan region and the Western Cape would probably receive nothing, whereas Zululand and the Eastern Cape would be in line for equalisation grants. Within the Central Witwatersrand metropolitan region, Johannesburg's Northern suburbs and CBD would not be entitled to subsidies from the metropolitan government, whereas community governments in Soweto and Alexandria would.
Frances Kendall, 57 Glenrose Road, Melrose 2196, Johannesburg, South Africa (tel 27 11 442 8898; fax 27 11 442 7247).