The Financial and Fiscal Commission (FFC) argues that indicators of ‘relative need’ are inadequate for the purposes of the horizontal split because:

• they don’t reveal whether provinces are adequately funded in the first place;
• they don’t take account of the variable costs that provinces incur in tending to these needs. These variations are due to factors that are beyond the immediate control of provincial administrations. The FFC argues that provinces should not bear the consequences of such unavoidable cost variations;
• they don’t measure the adequacy of the component shares. The size of the components in the current Department of Finance allocation formula is simply based on actual expenditure patterns.

In response the FFC has proposed a Costed Norms Approach (CNA) that:

• costs service delivery mandates. This would determine whether provinces are adequately funded to provide the stipulated basic services as required by the relevant national norms and standards;
• determines the relative sizes of each of the components on the basis of unavoidable costs instead of actual expenditure;
• estimates the variable provincial costs in service delivery (eg. higher costs in educating poor, rural and disabled learners).

The Department of Finance has responded by agreeing with the CNA in principle but arguing that:

• there is insufficient information available to implement this information-intensive approach to the horizontal division of revenue. In the absence of sufficient public sector performance data, there is no information about the links between the inputs and outputs in social service delivery. In the absence of such data, the FFC has recommended that input standards be costed. Even for such ‘input standards’ insufficient information would seem to be available, especially in the health and education sectors.
• Not all social service norms are sufficiently well defined. In the absence of uniformly clear input norms, it is unclear how the relative sizes of the social service components in the formula will be determined. Will a sector get less simply because it has no clearly defined norms? The Department of Finance has indirectly addressed the problem of component sizes by the work that it has done for benchmark budgeting in the social services. It has modeled the unavoidable social security and personnel expenditure in provincial budgets. In this way the integrity of provincial social service budgets has been improved. Given recent improvements in expenditure control and monitoring, it would therefore seem to be acceptable to base component shares on actual expenditure trends.

On reflection a number of further problems emerge:

• The CNA increases the likelihood that provinces will attempt to exclude ‘high-cost’ population from basic social service delivery. Given the time-lag in information generation, a province would be exposed to unfunded costs if ‘high-cost’ people moved into its jurisdiction. Only once data reflect the presence of such people in the province would the province be funded to deliver services to them. This problem is also present in the current Department of Finance formula. However, the FFC's closer definition of cost drivers has the potential to aggravate this dynamic.
• The FFC formula has clear benefits in terms of equity. It will enable better targeting of social expenditure to poor provinces. It does not, however, address the issue of whether these provinces have the capacity to spend the funds thus allocated.
• It has also often been argued that the CNA sets up an incentive to provinces to increase their costs in service delivery since this would give them access to additional funds in subsequent years. The FFC proposes to avoid this problem by only taking account of unavoidable costs. Inefficient and wasteful expenditure would not therefore be a basis for provinces to obtain additional funds. It is, however, not clear how a consistent distinction would be drawn between avoidable and unavoidable expenditure (for example, above average educator remuneration costs).

For the reasons stated above the acceptance and full implementation of the current FFC recommendations seems to be unlikely. This is supported by the fact that in April of this year the Budget Council recommended that the basic structure of the Equitable Share formula be retained until 2004. We interpret this to mean that the general approach (‘relative needs’) will not be departed from. In the short term it seems more likely that aspects of the FFC recommendations will be incorporated into the existing formula. A number of refinements could be made to the Department of Finance formula on the basis of what the FFC has recommended, without changing the basic structure of the formula. In what follows we compare the Department of Finance and FFC proposals for each of the four sectors and show where the FFC recommendations could enhance the existing formula.

1. Welfare

The existing Department of Finance Equitable Share formula distributes the welfare component on the basis of the target population in the Old Age Pension, the Child Support Grant and the Disability Grant. These target populations are weighted for the distribution of expenditure on each grant type. The impact of the means test on provincial take-up is reflected by an income adjustment based on the provincial share of the poor population. The Department of Finance formula does not have an explicit component taking account of variable phase-in rates for each of the grants.


The FFC bases its welfare component on the potential number of individuals eligible for each one of the six social security grants within each province. These numbers are adjusted for variable take-up rates by means of a phase-in parameter. The resulting numbers are multiplied by the average grant size for each of the grants. The FFC proposes that 5% be added to the total amount thus allocated to each province in order to account for Administrative costs incurred in providing social security. As is the case with the current Equitable Share formula, the FFC proposals make no provision for social welfare services. This omission is particularly disappointing given the Department of Welfare’s emphasis on developmental social welfare. The FFC cites the absence of norms and standards for social welfare delivery as reason for this omission. However, the Department of Welfare’s White Paper did recommend that provinces spend at least 20% of their budgets on this function. This provides a serviceable input norm for social welfare provision.

The Department of Finance allocates the welfare component on the basis of the need of recipients of three grants, whereas the FFC takes six grants into account. This refinement allows a more accurate distribution of revenue. Especially the inclusion of the Foster Care grant in the calculation of the welfare component is to commended since the demand for this grant is likely to increase in line with the increase in HIV/AIDS-related deaths. In this way the FFC proposal will be able to reflect new demands on provincial service delivery in a way that the current formula will not. Since the data are available to incorporate this innovation into the existing Equitable Share formula, we recommend that this innovation be implemented.

The FFC’s proposed 5% allocation for administrative costs is also an improvement on the existing formula. It should assist poorer provinces in eliminating administrative backlogs and increase provinces’ ability to deliver on their mandate. The slow take-up rate of social security in many of the provinces is largely due to inadequate administrative capacity. We therefore also support this FFC proposal.

2. Education

The Department of Finance’s Equitable Share formula allocates the education component of the Equitable Share on the basis of actual enrolments and the distribution of children between the ages of 7 and 18. The latter component is double weighted which means that inappropriate age learners are only counted as one third of an appropriate age learner. This mechanism is intended to encourage provinces to reduce the number of under- and over-age learners in the system. The latter group has been particularly connected with inefficiencies in the system.

The FFC argues that the amount necessary to meet a given educational standard will vary both across and within provinces, for reasons beyond the control of provincial education departments. This is based on the premise that learners from disadvantaged backgrounds (defined in terms of parental income and ruralness) require more educational resources to achieve basic educational outcomes. In calculating the average costs per learner, two factors are taken into account, namely, income level of families and the urban/rural distinction. Through various weightings, poorer rural and urban learners are allocated more resources in the education component of a province's equitable share. The FFC also proposes a distinction between public ordinary learners and special school learners, citing research that indicates that educating this group of learners is most expensive compared to all the other learner groups. In this way the proposed FFC education formula posits nine learner groups. This formula departs from the Department of Finance’s formula in that finer distinctions between different categories of learners are recognised.

One strength of the costing of basic education means that provincial education departments know exactly where the funds are derived from and who are targeted as the prime recipients of funding. This should reduce confusions about what the ‘core’ business of an education department is. The FFC formula also has the merit of distinguishing far better between the different kind of learner groups and their needs. In contrast to the existing education component, the FFC formula prioritises redress and equity in determining the education component of the equitable share.

The FFC formula only makes provision for what it calls ‘constitutionally mandated basic social services’. In the absence of national norms and standards about what basic educational services are, the FFC interprets it to mean primary, secondary and special school education. Educational programmes such as Early Childhood Development (ECD) and Adult Basic Education and Training (ABET) are interpreted as not falling under basic education and the FFC advises that these programmes should be catered for in the ‘basic’ component of the equitable share. This is so in spite of the fact that section 29 (1)(a) of the Constitution recognises ABET as a basic service in education.

The two formulas are very similar in their treatment of inappropriate age learners. The only difference is that while the existing education specifies the weightings of inappropriate age learners, the FFC leaves this open as a policy issue to be decided by central government.

Both formulas leave the division of the budget into personnel and non-personnel largely in the hands of provincial education departments (PEDs). Although the Norms and Standards offer guidance on this issue, its efficacy has been sorely tested in the recent history of provincial education spending. The fact that no separate amount is estimated for non-personnel expenditure probably means that PEDs remain responsible for this division ostensibly guided by the Norms and Standards.

The education formula should basically accomplish two things. In the first instance it should be equitable. One also expects the formula to have built-in mechanisms to encourage efficient spending in a PED. With this in mind we propose that:

• the method employed by the FFC in determining the provincial shares of the education component be accepted. The merit is that this formula provides a more refined indication of the need for education services in the provinces. As a result allocations in terms of the education component would be more equitable than they are in the existing Equitable Share formula;
• In order to encourage efficiency gains the existing formula for inappropriate learners should be retained. The FFC does not change the logic of the old formula here, although it leaves the question of the appropriate weights to central government.

1. Health

The total provincial population forms the basis of the health component of the Department of Finance’s Equitable Share formula. Because people without medical aid are more likely to use public facilities, they are weighted four times higher than people with medical aid. In 1998 the non-medical aid users were also weighted for differences in utilisation rates due to age and gender. This weighting was not carried though in subsequent years.

The FFC bases its health component on the total provincial populations weighted for differential utilisation rates. It does this by weighting for age and gender, and then adjusting for relative poverty. To this weighted population the FFC then applies a per capita annual cost of primary health care provision (direct and indirect service costs estimated). The indirect services charge is also adjusted for population density, thus benefiting provinces with a low population density, where it would be relatively more expensive to provide health services than in more densely populated areas. The FFC proposes that provision for secondary health services be included in the proposed enlarged basic component of the Equitable Share formula.

The first important difference between the two approaches is that the FFC adjusts for differential utilisation rates, while Department of Finance does not. These utilisation rates are important because they provide a refinement on the incumbent formula’s indicator of relative need. Health services in provinces where large parts of the population are poor, female or elderly would be in greater demand than in provinces where this is not the case – even if these provinces had exactly the same size total population. This refinement could easily be incorporated into the existing Equitable Share formula and would make it more responsive to the needs of the most vulnerable members of our society.

The second important difference is the adjustment for the impact of population density on service delivery costs proposed in the FFC formula. This adjustment is likely to benefit provinces like the Northern Cape that are less likely to benefit from the economies of scale available to provinces with large metropolitan areas. Demographic and geographic variables are beyond the control of provinces, and they should therefore not be disadvantaged as a result of such factors.


The third big difference between the two formulas is that the FFC uses a poverty factor and not the medical schemes factor to adjust for utilisation rates. Even though there is a high degree of correlation between poverty and people without medical aid, poverty is chosen as adjustment because many people without medical aid will still use private facilities and vice versa. The FFC thus judges that poverty is a better indicator of utilisation rates than the number of medical aid holders. In any event the benefit from this adjustment seems to be marginal, depending on where the poverty parameter is set.


A fourth difference is that the FFC proposed to divide health resources on the basis of primary health care costs alone. The current Equitable Share formula tries to establish relative need for all provincial health services. An earlier FFC proposal to include secondary services has been excluded from the final recommendations. From what the FFC proposes it is not clear what the implications of this changed position would be. It however appears to hold negative consequences for provinces with well-developed secondary services such as the Western Cape and Gauteng.

2. Other components

The non-social service components of the Equitable Share formula are intended to fund non-social responsibilities of provinces. In the Department of Finance’s Equitable Share formula these components are residual. While the size of the social service components are determined by actual expenditure trends in the provinces, the remaining components are made up of whatever is left over after the size of the social service components have been determined. This is borne out by the fact that not one of the social service components have decreased in size since 1998, while the size of all the non-social service components has decreased at least once.

The FFC proposes that the relative size of these components be linked to the size of the basic social service components. This would mean that the appropriate non-social service funding for a given level of social service funding would be considered, and not be merely residual. This approach should be supported in principle, but it becomes difficult to implement given the problems with the absence of information and clearly defined norms outlined above.

The FFC also recommends that the existing backlog and economic activity components be consolidated into one basic component that would fund all remaining social and non-social services. The FFC does not explain why the economic activity share should be scrapped, but it does point out that the backlog component is not well targeted and that this policy goal should rather be funded by a conditional grant. Currently social service pressures on the provinces often swallow up the funds allocated via the backlog component.

We recommend that the Economic activity share be retained in order to fund provinces for the costs of economic growth. Provinces that have the capacity to support national growth and provide employment should be supported to this end. At the same time provinces that do not currently have such capacity should not be prevented from developing it. The backlog and basic shares could play this role. We therefore recommend that the basic component be incorporated into the backlog component to further support poor provinces in developing capacity in capital and current expenditure backlogs. This would restore the pro-poor bias to the basic component that it had prior to 1999.

Such an approach would go some way towards formulating a coherent approach to the ‘spending capacity’ problem often highlighted by Department of Finance officials. They argue that they do not want to allocate additional funds to provinces that are incapable of spending such funds. While such a concern has merit, it should not be used an excuse to not fund poorer provinces to develop such capacity. The approach formulated above would reward effective provinces at the same time as developing capacity in provinces where such capacity is absent.

The institutional component in the Department of Finance formula is intended to fund the costs of running a government and providing services that are not directly related to the size of a province’s population. In the current formula it comprises 5% of the total Equitable Share. The FFC has recommended that this component be reduced to only fund a basic governmental structure. The FFC does not provide any rationale for this proposed change, and it is therefore difficult to assess its value.

2. Conclusion

We argued above that the FFC recommendations should not be accepted as a whole. Specific recommendations should, however, be incorporated into the existing Department of Finance formula. The following adjustments could be made to the existing Equitable Share formula without requiring the generation of additional data. We believe that such adjustments would render the existing formula more equitable.

• The FFC proposal for the education component provides a more refined indication of the need for education services in the provinces than the current Equitable Share formula does. The FFC proposes to base the distribution of the education component on nine different learner groups, each with its own needs and cost implications. We support the attempt to cost the education share in such a re-distributive manner.
• The FFC's welfare component is constructed along the same lines as the education formula in that there is a clear attempt to sufficiently distinguish the target recipients of the various social security grants. The Department of Finance allocates the welfare component on the basis of the need of recipients of three grants, whereas the FFC takes six grants into account. In addition the FFC proposes that 5% must be added to the total amount allocated to each province’s administration costs in provincial social security. We support this recommendation because we believe it will go some way in reducing the administrative backlogs and enable provinces to deliver on their mandates. The one disappointing feature of both proposals is that they do not make explicit provision for social welfare.
• The proposed health component also holds improved targeting by suggesting that variable utilisation rates based on age, gender and poverty. Such a formula could improve on the distribution of incumbent formula that merely looks at Medical Aid holders as an indicator of likely utilisation rates. We also support the FFC proposal that health allocations be adjusted for the impact of population density on service delivery.
• We support the FFC proposal that the basic component be incorporated into the backlog component. This would restore the pro-poor bias to the basic component that it had prior to 1999.