COSATU on Sugar Tax Part 1 of 3
COSATU on Sugar Tax Part 1 of 3
Interview with Sdumo Dlamini on unity and cohesion of COSATU
Talking NHI with Lebo Mulaisi
Subscribe to Cosatu Whatsapp

The Shopsteward Subscribe to get a copy of the Shopsteward The Shopsteward Online Archive

Shopsteward Volume 26 No. 2

COSATU Media Monitor COSATU Media Monitor COSATU Media Monitor


Tel: (011) 339-4911
Fax: (011) 339-5080/339-6940
Email: donald @ cosatu . org . za

For comments on the website email: donald@cosatu.org.za

Media Centre  |  Proposals

The People's Budget 2001

Proposals by COSATU, SACC, SANGOCO, 21 February 2001


The People's Budget 2001: Summary


Table of Contents

  1. Introduction
  1. What is the People's Budget?
  2. Aims of the People’s Budget
  1. The challenge of reconstruction and development
  1. Poverty and inequality
  2. Factors leading to poverty and inequality
  1. Toward a developmental budget
  1. The developmental role of the state
  2. Fiscal implications of the proposed growth path
  1. Translating Vision into Reality
  1. A developmental fiscal package
  2. Building Public Service Capacity
  3. Towards Solutions

Appendix 1


Box 1: What’s Happening with Unemployment?

Box 2: Preventing Mother-to-Child Transmission of HIV

Box 3: A Framework for Assessing Public Spending Acknowledgements

  1. Introduction

This document presents an alternative to the current macro-economic framework, one that meets the aspirations of a nation aiming to accelerate social transformation. Our proposals should contribute to a shared national vision on growth and development. They show that a better way to meeting our developmental challenges is not only possible, but also a national imperative.

These proposals, developed jointly by the Congress of South African Trade Unions (COSATU), South African Council of Churches (SACC) and the South African NGO Coalition (SANGOCO), move beyond criticism of existing policies to provide workable, credible and progressive alternatives.

We look forward to discussing our ideas with communities, government, parliament and business. We pose to all our partners the challenge of going beyond criticism to the development of constructive proposals.

This document has four sections:

  • A brief overview of the People’s Budget Campaign
  • An assessment of the development challenges facing government and the unintended consequences of the current macro-economic policy
  • An integrated set of proposals for improving government programmes
  • Measures to ensure that those proposals are carried out, including a developmental approach to fiscal policy and ways to enhance the capacity of government to deliver.

1.1 What is the People's Budget?

The People's Budget is an initiative of Congress of South African Trade unions (COSATU), South African Council of Churches (SACC) and South African National NGO Coalition (SANGOCO), with technical support from the National Labour and Economic Development Institute (NALEDI).

The People’s Budget aims, first and foremost, to stimulate broad discussion and mobilisation around how government uses its resources. To that end, it will produce a fiscal framework every year, based on in-depth research. The framework will explore the main tasks facing government in ensuring social and economic development and growth, and assess the fiscal implications.

The People’s Budget will not attempt to provide line-by-line alternatives to the national budget. We are not trying to replicate the work of government. Rather, we aim to demonstrate where strategic changes in fiscal policy and expenditure patterns can intensify the fight against poverty and ensure sustainable economic development.

This year’s framework is based on:

  • Budget hearings with civil society formations;
  • Consultation with a variety of organisations and structures, both inside and outside the SACC, SANGOCO and COSATU; and
  • Research into fiscal and economic issues, in part by NALEDI, and the Economic Policy Research Institute (EPRI).

In the coming year, we will continue with research into the budget itself, focusing on provision for land reform, HIV/AIDS, free basic services and skills delivery; and on ways to improve government spending, including a deeper analysis of under-spending, the restructuring of state-owned assets, public-private partnerships and prescribed assets.

The People’s Budget will also develop educational material and campaigns to ensure broad mobilisation around the budget and to engage with the Treasury, Cabinet, Parliament, and the wider South African public.

1.2 Aims of the People’s Budget

The People's Budget arose largely in response to the deep budget cuts in public spending since the introduction of GEAR in 1996. These cuts have reversed many of the gains we made in the first few years after the transition to democracy. The Speak Out on Poverty hearings convened by COSATU, SANGOCO and the SACC in 1997 helped to expose the impact of these cuts on people's lives.

The People's Budget seeks to demonstrate that the ideology of "There Is No Alternative" (TINA) is profoundly mistaken. At the same time, it will increase the capacity of civil society to influence the budget process, and support the constitutional right of parliament to amend the budget.

The People’s Budget will define fiscal strategies that can alleviate poverty, support economic development and ensure greater equity by race, gender and class, by

  1. meeting basic needs, especially by restoring and enhancing public services and social spending,
  2. ensuring the retention and creation of quality jobs in a context of economic growth,
  3. giving the majority of our people greater access to assets and skills,
  4. enhancing democratic and participatory governance, and
  5. protecting the environment and ensuring development throughout the southern African region.

  1. The challenge of reconstruction and development

The key problem facing South Africa remains the poverty and inequality left by apartheid. We here first review the extent of the problem, and then outline a development strategy to address it. Finally, we contrast that strategy with GEAR.

2.1 Poverty and inequality

A recent report by the United Nations Development Programme argues that:

The final fifteen years of the apartheid era saw a massive transfer of wealth from the poor to the rich: between 1975 and 1991, the income of the poorest sixty percent of the population dropped by about 35%. By 1996, the gulf between rich and poor had grown even larger. The poorest quintile received 1.5% of the total income, compared to 65 percent received by the richest quintile and 48% percent by the richest 10%. (United Nations Development Programme: 2000)

The report ranks South Africa as the third most unequal society in world, surpassed only by Brazil and Guatemala.

The massive inequalities in income and wealth translate into extensive poverty. About half of all South Africans, some 18 million people, live in households that earned less than R350 a month in the mid-1990s. Poverty is worst in the rural areas, as the following table indicates. The regions that include the former homeland areas have far higher rates of poverty. They account for over two thirds of all poor households in South Africa.

Table 1. Poverty by province


% of Population in Poverty

Provincial Share of National Poor

Northern Province



Eastern Cape









North West



Northern Cape



Free State






Western Cape



Source: UNDP 2000

2.2 Factors leading to poverty and inequality

South Africa faces a classical poverty trap. In this situation, massive inequalities and associated poverty prevent growth and development, which in turn worsens poverty. Specifically, poverty lowers the productivity of the labour force by making it harder for people to acquire skills and by undermining social cohesion. Since it reduces household incomes, it limits domestic markets. For this reason, the RDP argued that growth must be combined with redistribution.

The following diagram portrays the vicious cycle that underlies the poverty trap, where poverty reinforces low growth, which in turn worsens poverty.

South Africa the poverty trap

After the transition to democracy in 1994, the government responded to this vicious cycle with four strategies that, at least in part, contradicted each other:

  • A robust effort to redirect government spending toward historically deprived communities,
  • GEAR,
  • An industrial strategy focused primarily on improving competitiveness and especially exports, and
  • A labour policy designed to shift from the oppressive colonial relationships of apartheid to a modern labour market.

GEAR reinforced the vicious cycle of poverty by supporting an economic strategy that did little to support greater equality. Because its proposals for restructuring the economy remained weak, it effectively maintained South Africa’s historic growth trajectory. That growth path effectively emphasises minerals production and refining for export – which generates few jobs, strengthens big business, and reinforces the underdevelopment of the rural areas. Meanwhile, tight money policies were enforced through measures to increase interest rates, discouraging domestic investment.

At the same time, GEAR called for measures to reduce the bargaining power of labour. As that would limit both wages and skills development, as well as aggravating conflict in the workplace, it would worsen productivity problems. This component of GEAR contradicted the new labour policy, which involved new laws designed to encourage the more productive use of labour through greater permanency, protection of labour rights and collective bargaining, and widespread skills development.

For the budget, a central contradiction emerged between the need to improve government services for the majority of South Africans and the tight restrictions placed on government expenditure by GEAR’s fiscal targets. As a result, as discussed in Chapter 3, spending on social services and infrastructure declined steadily in real terms in the second half of the ‘90s, undermining a central strategy for addressing poverty.

In the event, the failure of the GEAR strategy emerged in massive job losses, slow growth and low investment in the past three years. The following table contrasts real achievements with the GEAR targets. It shows that government cut the deficit even more than expected, but the outcomes in terms of development and economic growth did not come close to expectations. In particular, employment fell massively, instead of increasing; private-sector investment grew only painfully slowly; and GDP growth averaged 2,4 per cent a year, compared to GEAR’s projection of 4,2 per cent.

Table 2. GEAR projections and actual achievements, 1996-‘99


Annual average, 1996-‘99


Projected in GEAR



Fiscal deficit as percentage of GDP



Average tariff as % of imports



Real bank ratea




Annual change in formal, non-agricultural employmentc

270 000

-125 200

Real private sector investment growth



Real non-gold export growthb



GDP growth



Inflation (CPI)



Sources: South African Reserve Bank, Quarterly Bulletin, June 2000; Department of Finance, Budget Review 2000; Department of Trade and Industry, Economics Database.
Notes: a. for actuals, residential bond rate less CPI. b. for actuals, real non-mining export growth. c. figures for 1996 to 2000.


Box 1. What is happening with employment?

Government spokespeople now argue that employment increased significantly after 1998, based on figures provided in the 1999 October Household Survey (OHS), published by Statistics South Africa. They suggest that employment in the informal sector and agriculture has recently grown at least as rapidly as the formal sector shrank.

Unfortunately, the 1999 OHS data on employment outside the formal sector do not correlate with those of the 1998 OHS, making it impossible to define trends. In effect, both give a snapshot of employment using different coloured lenses – probably both reasonably accurate, give their divergent definitions and methods, but not useful in defining changes over time.

The 1999 October Household Survey shows substantially higher employment than in 1998. As a result, it suggests a small decline in the unemployment rate – from 37,5 to 36,2% under the expanded definition, and from 25,2 to 23,3% under the narrow definition that counts only those actively seeking work.

These data contradict other trends in the economy. They suggest employment increased by 10% in 1998-’99, even as the economy grew only 0,8%. That would mean a massive reverse in the trend toward declining labour absorption with economic growth over the past two decades.

Sectoral figures are even more problematic. The bulk of the jobs supposedly created – some 770 000 – are in the informal sector. If this reflected growth, rather than a change in the definition of employment, the sector would have grown at the extraordinary rate of 50% in a single year.

Similarly, the data show growth of 17,5% in agricultural employment, while output rose only 4% in that year. The survey admits the change may reflect the fact that this year, for some undisclosed reason, some subsistence and informal producers in the rural areas were classified employed, where in the past they were recorded as unemployed or economically inactive.

In any case, the OHS figures count any income-generating activity under informal employment, including parking cars on the street and hawking vegetables.1 These activities do not provide a livelihood or raise productivity. They are survival strategies, sometimes only disguised unemployment2, that do not address the economic and social consequences of job losses and poverty.

  1. Toward a developmental budget

The diagram below depicts how a more developmental strategy could break the vicious cycle of poverty, building on the reinforcing nature of the component macro-economic policies. This strategy bolsters the effectiveness of fiscal policy, which will have a greater effect promoting economic growth and job creation if households have higher living standards and greater access to resources and skills. Higher rates of economic growth also expand the fiscal resource base, supporting even greater levels of social delivery, accelerating the virtuous cycle.

An alternative macroeconomic strategy

To break out of the poverty trap requires an unambiguous conceptualisation of the strategic role of the state in development. The budget must then be reshaped to accord with that view. For this reason, the following section outlines the developmental role of the state. It then considers the implications of the strategy for government expenditure.

3.1 The developmental role of the state

In addition to the normal administrative functions, an effective development strategy requires that the state fulfil four roles.

First, the state must drive a growth strategy that focuses on providing strong policy support for sectors to protect and create quality jobs, meet basic needs for the poor, and expand exports.

Second, the state must provide a social wage that sets a floor under living standards for all South Africans. The social wage comprises government services and grants provided to households in addition to earned income. It should ensure that no one faces absolute poverty.

The social wage must

  • provide welfare grants, education, health care, policing and housing at a level sufficient to support community development;
  • accelerate skills development and improve education on a mass scale;
  • develop more efficient, socialised systems to meet retirement, transport and health needs for working people, which effectively enhances the efficiency of the economy as a whole; and
  • be designed to support economic growth and employment creation.

Third, the state must also transform itself to strengthen democracy and the public sector through the establishment of systems to permit greater participation by the majority, who historically have been shut out of power, as well as more coherent and effective delivery systems and structures. Steps to achieve this end include establishment of participatory procedures for policy development and strengthening Parliament, especially to amend money bills and participate in policy development. They must also include measures to control lobbying by big business and to limit patronage and corruption.

Finally, the developmental state must support alternative centres of economic power. All the other strategies can contribute to this aim. They can change the nature of wealth by supporting a stronger state sector, co-operatives and small and micro enterprise, linked in part to land reform, improved housing, and investment in skills development.

3.2 Fiscal implications of the proposed growth path

Adapting the budget to overcome the poverty trap requires that the budget increase to support

  • an effective strategy to restructure the economy, co-ordinated with measures to improve the social wage;
  • expansion of the social wage in general, with the introduction of a Basic Income Grant, and establishment of National Health Insurance, and stronger measures to combat HIV/AIDS;
  • a coherent skills development strategy; and
  • better integration of infrastructure and housing development with the overall growth strategy.

We here first consider the proposals for achieving these aims. Section Four then explores the broader policy changes required to implement them – above all, a developmental fiscal strategy and increasing the capacity of the state.

3.2.1 Fiscal implications of economic reconstruction and development

To create jobs and establish a more stable basis for growth, a growth strategy must systematically restructure production. In particular, the economy must move away from the more traditional export sectors of minerals refining, which are associated with limited employment creation and large-scale capital, to relatively labour-intensive industries. The strategy should focus, not only on expanding exports, but also on ensuring production of goods that meet basic needs, reducing the cost of living while raising living standards. Growth sectors would then include, for instance, agriculture and food processing; services and tourism; infrastructure provision; and cultural industries.3

A critical first step is to develop sectoral strategies to guide government, capital and labour. These strategies must ensure that production shifts to meet changes in domestic demand that result from improvements in the social wage. They must also help diversify exports. Government can support sectoral strategies through:

  • Targeted investment in transport, communications, schools, clinics and other infrastructure;
  • Measures to ensure that the financial sector supports new kinds of production, including self-employment and co-operatives, as well as low-income housing;
  • A review of tariffs to ensure that reductions do not unnecessarily harm employment;
  • Rigorous implementation of the skills development process;
  • Greater support for constructive implementation of new labour policies including in agriculture, domestic labour and other sectors that must make a substantial adjustment;
  • A strategy to ensure retail networks support domestic production and ensure efficient distribution of wage goods; and
  • Rapid redistribution of land and measures to ensure the security of tenure.

In this context, the expansion of the social wage must be shaped to stimulate investment and employment. For instance, housing programmes must locate workers near to jobs, to minimise commuting costs; education must provide necessary skills; and socialised health care must control the soaring cost of medical aids.

The main new costs of the proposed growth strategy arise from improvements in the social wage and skills development, which are discussed below, and from land reform.

3.2.2 Improving the social wage

The apartheid system left a heavily distorted structure of spending on the social wage, with very high levels for historically white communities and virtually no support for black households. Overcoming the poverty trap requires that this situation be overcome urgently by extending an adequate social wage to all communities.

As the following table shows, discrimination in government spending between communities under apartheid resulted in huge differences in infrastructure by province. The provinces that incorporated substantial homelands show the largest infrastructure deficits. Similar differentials exist in the main social services – health, education, welfare and policing.

Table 3. Infrastructure provision by province, 1996


% of households without basic:


Thousands of households




Refuse removal


Western Cape, Gauteng







Northern Cape, Free State, Mpuma-langa, North West, KwaZulu Natal







Northern Province, Eastern Cape







Source: Calculated from, Department of Finance, Budget Review 2000

Between 1994 and 1996, substantial progress was made in improving social services and infrastructure in historically black communities. Thereafter, the restrictive fiscal policy associated with GEAR led to fresh cuts in the social wage. In 1996-‘99, real expenditure on education, health, welfare and infrastructure fell each year, while police and housing remained virtually unchanged. The decline in per person terms was obviously even larger, given population growth of between 5 and 7 per cent over the period. As a result, the provision of new infrastructure and housing remained slow, with a particular poor record for maintenance. The main social services experienced stagnant delivery at best.

Table 4. Changes in expenditure on social wage by function, 1996/7 to 1999/2000, in nominal and real terms (deflated using CPI)

Expenditure in millions of rand

Average annual change

% of non-interest spending,





after inflation


























Transport & communication


















social wage functions


















Source: calculated from, Department of Finance, Budget Review 2000, using CPI figures supplied by Statistics South Africa.

Declining real spending on the social wage undermined productivity growth, underpinning a low-wage labour market characterised not only by severe unemployment but also a large number of the working poor. At the same time, it further limited domestic demand.

The MTBPS expected to reverse the decline in the social-wage functions, permitting real growth of about 2 per cent overall. But the bulk of the increase will benefit defence, economic services and infrastructure. Education, health, welfare and policing still face reductions in spending per person over the next three years.

As the following table shows, in the next three years education will be the slowest-growing function. Yet the MTBPS itself points to the critical importance of improving skills to bring about growth and boost employment creation. In contrast, defence enjoys the third fastest rate of growth, considerably more rapid than health, education, welfare and the justice system. This prioritisation appears to ignore our greatest security risks – that is, rising unemployment and poverty.

Table 5. MTBPS proposals on spending by function in real terms, per capita

In bns of 1999 rand*

Average annual growth





Per capita**

Economic services










Defence and intelligence















Integrated justice system




















*deflated using CPI. ** estimated at 2,3 per cent a year.
Source: Department of Finance, MTBPS 2000

Overcoming the poverty trap requires that the real annual increase in expenditure on infrastructure, welfare, health, the justice system and education be at least 2 per cent over population growth for the next four years. By 2004, that will restore spending, in real per capita terms, to 1996 levels.

To accommodate this increase, by 2003/4 the budget as a whole will have to increase by R12 billion over current MTBPS projections. The overall rate of growth of the budget would then climb to 4 per cent a year in real terms.

Within the overall increase in spending, the People’s Budget proposes the following programmes to strengthen the social wage.

3.2.3 The Basic Income Grant

The system of social grants should ensure a minimum income for all South Africans. The current set up does not come close to reaching that goal. For this reason, People’s Budget proposes a basic income grant as a critical component in an effective comprehensive social security system. The basic income grant would provide an effective, administratively easy mechanism to substantially raise the income of poor households.

The current system of grants shows three major weaknesses.

First, research conducted shows that over 13,8 million people in the poorest 40% of South Africa’s households do not qualify for any social security transfers. (Haarman 1998) Only the old-age pension is adequately focused on poor households. The other grants are still disproportionately located in the Northern and Western Cape. Overall, welfare spending per person is not well related to the level of poverty in a province.

Second, old-age pensions and child-care grants account for the bulk of payments. These grants are important in alleviating hardships, but are targeted according to age and so may do not reach all poor households.

Third, most grants have declined in real terms in recent years, and none of them reach the entire population for which they are designed. Relative to headline inflation, the old age pension and disability grants have dropped by almost 20 per cent. The child grant has not increased since its introduction in 1998, so relative to inflation it has already fallen by 10 percent. Compared to the inflation rate for low-income people, which is somewhat higher than the overall CPI, these grants have fallen even more.

The latest MTBPS proposes that welfare spending as a whole should grow at 2,2 per cent, just below the rate of growth in the population. Within that amount, however, the Welfare Department proposes to cut the share of spending on social security from 90 to 80 per cent of the budget. This will cause further restrictions on grants.

In sum, the current system of grants represents a critical support for many of the very poor. But measured against the needs of the poor, it remains inadequate.

In light of the difficulties with the existing social grant system and the delays in extending basic services to the poor, the People’s Budget calls for the introduction of a basic income grant. The basic income grant would involve a relatively small sum – between R100 and R200 – paid to all individuals every month. Because it is a flat sum, the proposed grant is highly progressive.

Obviously, the central aim of the proposed grant is to increase incomes for the poor. But the grant would also provide a degree of household and community development, laying the foundation for more productive and skilled communities over time. It would ensure that even the poorest households have a little cash to support economic engagement.

Universal coverage and sustainability are crucial for the basic income grant to take effect as proposed.

Universal coverage – payment to all individuals – will:

  1. Minimise administrative burdens. As the current systems for social grants and municipal services demonstrate, government simply does not have the capacity to utilise means tests efficiently. The welfare administration is already over-stretched by the existing grants. Payment through the PostBank could further streamline delivery.
  2. Favour larger households, which tend to be poorer and have greater needs.
  3. Benefit women and young people directly. That should increase their power and autonomy within the household economy, which in the long run should help reduce the abuse of women and children. That in turn should support social cohesion and skill enhancement, limit the spread of HIV, and ultimately sustain social and economic development.
  4. Reach not only the unemployed but also the working poor, who constitute a very large share of people officially classified as employed. Farm and domestic workers, people trying to eke out a living in the informal sector, and even large sections of the formal sector earn incomes well below the poverty line.
  5. Reduce the disincentive to work and stigmatisation associated with the dole as found in other countries, where a means test is applied. In effect, a means test may require people to choose between earned income and the grant.
  6. Help overcome the risk of corruption, by reducing officials’ discretion over who gets the grant.

A second critical aspect is fiscal and economic sustainability. This sustainability has two foundations: the proposed financing mechanisms, and the longer-term stimulus to the economy.

Essentially, the grant would be partially funded through a solidarity levy on the high-income group. The levy would recoup the cost of the grant to the well-off, plus some additional funds from the very high-income group. We propose a levy in the form of a surcharge on income tax averaging 17,5 per cent for the top two quintiles. EPRI studies of tax effort indicate that this levy is sustainable. Haarman (2000) calculates that the levy would raise around R15 billion, covering three quarters of the cost of the grant. It would be borne essentially by the top quintile of income earners.

More fundamentally, the sustainability of the grant must be seen in a dynamic context. As the grant provides the basis for economic growth in the medium term, the fiscal burden reduces over time relative to the budget and the economy as a whole. In that sense, it represents a critical investment, whose longer-term returns more than justify the short-run costs.

In sum, the basic income grant essentially ensures an adequate social wage for all the poor by overcoming the weaknesses in public administration and the inappropriate financing systems that bedevil current systems. In the process, it should support social and economic changes that are critical to overcoming the poverty trap and initiating sustainable development. Moreover, it will give a degree of stability to households, protecting them from the vagaries of the market.

If a grant of R100 a month were provided to all South Africans aged over 17 and under 65, the cost would come to around R25 billion a year. If paid to all individuals over six, the cost would come to about R40 billion a year. About R15 billion of the cost would be met through the solidarity levy.4

3.2.4 National Health Insurance

South Africa spends 8,5 per cent of its GDP on health, yet the outcomes – measured for instance by child, infant and maternal mortality and life expectancy – are worse than other middle-income developing countries. The reasons lie in part in inequalities in the health system, and in part in the lack of basic infrastructure to ensure clean water and sanitation for many of our people. The HIV pandemic has aggravated the situation. To remedy this situation, the People’s Budget proposes a national health-insurance scheme.

The co-existence of private and public health care forms an important factor behind the inefficiency of health spending in South Africa. The private sector provides the best care in the world for those who can afford it, while the majority continues to rely on inadequate public services. Private health care serves less than 20 per cent of the population, but absorbs two thirds of total health spending. In contrast, the public health sector serves the majority of our people, but remains under-resourced, resulting in poor quality care.

Only about 7 million South Africans, of whom only 9 percent are Africans, are covered by private medical schemes. The annual revenue of the private medical schemes comes to R25 billion a year, about 20 per cent more than the total public health budget.

Meanwhile, the public health system faces rising demand and falling budgets. After 1994, government opened facilities to all races, provided health care free to mothers and young children, and expanded primary health care. These measures should ultimately reduce costs for curative care, but initially led to higher referrals to hospitals, with a consequent increase in costs there as well.

In the same period, budgets for health declined relative to inflation. Between 1996/7 and 1999/2000, provincial health budgets rose 4 per cent a year, with inflation at 7,7 per cent a year. In other words, in real terms health budgets dropped by over 3,5 per cent a year, or close to 10 per cent in three years. At the same time, the population was growing at over 2 per cent a year. In per capita terms, then, the health budget dropped by 15 per cent over the period. Of the new health care facilities constructed in the past six years, 46 per cent5 did not have an adequate supply of clean water and electricity. Many clinics are opened without necessary staff or resources.6

The two-tier system promotes a mal-distribution of resources and wastage, inflates health costs, and defeats the commitment to a health care for all. On the one hand, private health care has seen extraordinarily rapid increases in tariffs, largely due to excessive profits and administration costs, with increased instability among medical-aid schemes. On the other hand, the private health care system actively weakens the public health care system by shifting the cost of caring for patients with serious illnesses, including HIV, onto the public sector; and by setting up a parallel system of (expensive) private hospitals that diverts potential paying patients from public hospitals.

The People’s Budget proposes National Health Insurance (NHI) to end the two-tier system by incorporating all health resources into the public sector. A new NHI Authority would allocate the health budget to hospitals and practitioners. It would be funded by the existing budget plus a progressive dedicated levy equal to existing private health costs.

The levy would be on high incomes, both salaries and other, and would effectively replace the current cost of health insurance and medical schemes. Because it would replace employee-employer insurance premiums and out-of-pocket expenditures, it should not increase the cost of health care to the South African with medical aid or society as a whole.

In the long term, by reducing administration and procurement expenses, the cost of health care should ultimately decline. Instead of confused and often unjust dictates of insurance companies, a greatly expanded programme of technological assessment and cost-effectiveness would guide decisions about covered services, as well as about the allocation of funds for capital spending, drug formularies and other issues.

Each hospital would receive an annual global budget to cover all operating expenses. For patients not using hospitals, the diversity of existing practice arrangements necessitates a pluralistic approach. To minimise disruption, the NHI would include three payment options for doctors and other practitioners: fee-for-service payments, salaried positions in the institutions receiving global budgets, and salaried positions within group practices receiving per capita (capitation) payments.

The NHI would not increase the burden of health costs on society in the short run, and should reduce the burden per person in the longer term. That has great importance for controlling the cost of living and of production. It would, however, lead to an apparent increase in the budget, equal to the amount now spent on private health care.

The structure and financing of the NHI is meant to be self-funding, as it will be raised as a levy. The funds raised through the levy would be ring-fenced, and used for medical care for all South Africans.

3.2.5 Treatment and prevention of HIV

Undoubtedly, HIV/AIDS forms on the most important daunting health problems facing South Africa, with the rate of infection currently estimated at around 25 per cent for adults. The budget must begin to ensure that people with HIV have adequate access to social security and treatment, and that condoms are available to all adults.

The epidemic also has a direct impact on the social security system, since most HIV-positive people require state assistance. Only a relatively small percentage of people living with HIV/AIDS have medical cover, retirement or unemployment support, and few benefit from existing social grants. People who live with AIDS often face discrimination from officials when trying to access government services like disability grants.

A number of existing grants can be targeted to support people living with HIV/AIDS, including child support, foster care and disability grants. Rather than establishing a separate grant for people living with HIV/AIDS, government should adopt a conscious strategy of directing existing grants to meet their needs.

A key element in combating HIV/AIDS is effective treatment strategies. Currently there are no effective strategies for treatment in the public sector. Debate has centred on the cost of drugs and the effectiveness of anti-retroviral medicines. Critical treatment strategies include obtaining cheaper medication; treatment for opportunistic diseases and STDs; use of anti-retroviral therapies including, but not restricted to, prevention of mother-to-child-transmission; and establishing a proper infrastructure for counselling.

It is also critical that government make condoms available on a broader basis. The current budget for condom provision suffices to provide around five condoms per adult South African a year. Yet many adults cannot afford to buy condoms from the private sector. Moreover, in remote rural areas they may not even be available for sale. It is clearly crucial that the government address the shortfall.

We here only consider the fiscal implications only of anti-retroviral treatment for HIV-positive people. The cheapest anti-retroviral treatment consists of AZT with some additions; they are not as effective as the top-of-the-line combination, but will still prolong people’s lives and health substantially. In Brazil, which has provided low-end generic anti-retrovirals on a large scale, deaths are down by half, and the rate of infection has dropped.

Obviously, the use of anti-retrovirals would involve researching affordable options. News reports currently suggest that use of Indian generic medicines could reduce the cost to between US $ 350 and US $ 600 a year – at the current exchange rate, between R230 and R400 a month, or rather less than the old-age pension.

If the infection rate is around 20 per cent of the adult population, at R500 a month it would cost around R6 billion a year to give the cheapest anti-retroviral treatment to 20 per cent of those infected – which is probably the number with symptoms at any given time. This is probably not a high enough rate of treatment to limit the spread of infection, but would go far to prolonging lives until a better treatment is found.

In addition, to provide every South African adult with two condoms a week will cost around R1 billion, at R1 per condom. That sum should also cover distribution costs.

In short, just providing retrovirals and condoms on a near-adequate scale would cost approximately R7 billion. If not offset by other savings, that would increase the current public health budget by around a third. In fact, however, the combination of improved health for HIV positive people and reduced spread should lead to savings in treatments for opportunistic diseases, training and productivity. Furthermore, the entire cost of the proposed HIV treatment need not come entirely from the budget.

We need much more research to assess the net impact of the proposed measures on the budget. It seems likely that it would be no more than R5,5 billion a year – probably an overestimate of the costs in the medium-term.

Box 2. Preventing Mother-to-Child Transmission of HIV/AIDS

A study by Ned Geffen of the fiscal cost and the cost-effectiveness of using anti-retroviral drugs (Nevirapine – NVP – or AZT) to reduce mother-to-child transmission of HIV comes to the following conclusions.

There is a consensus in this literature that (1) a mother-to-child transmission prevention (mctcp) programme would, at worst, cost the state less than 3% of the total health budget and probably far less, that (2) an mtctp programme using the short-course AZT would be cost-effective, and that (3) a properly implemented programme would substantially reduce the number of HIV infections (estimates range from 11,500 to over 23,000).

Where the studies have taken into account the cost the state must incur for treating HIV+ children, there has been a consensus that not only would the short-course AZT regimen be cost-effective, but that it would be cost-saving.

This finding extends to an mtctp programme using NVP, since this would be a substantially cheaper option with similar health-benefits.

Importantly, the findings of Hensher (2000), a recent internal briefing commissioned by the Department of Health, concur with this consensus.

Two scenarios have been generated for calculating the cost and cost-effectiveness of an mtctp programme:

Scenario one: The programme uses the NVP regimen. All costs and other data employed in the calculation are based on current prices and the latest available data. Where there is a dispute between reliable sources, the estimate that reduces cost-effectiveness has been used. The cost of this scenario is less than R155 million per year which is less than 1% of the health budget. The cost-effectiveness of this scenario is approximately R365 per DALY, which is easily within the World Bank recommendation. Approximately 23,000 infections are avoided in this scenario.

Scenario two: The programme uses the short-course AZT regimen with the corresponding adherence rates, but in all other respects is identical to scenario one. The cost of this scenario is slightly less than R225 million per year (also less than 1% of the health budget). The cost-effectiveness of this scenario is approximately R630 per DALY, also comfortably within the World Bank recommendation. Approximately 19,000 infections are avoided.

Model results




HIV infections averted less deaths from infant formula milk

Nevirapine according to HIV 012 regimen


R344 per DALY


Short-Course AZT according to Thai-CDC regimen.


R606 per DALY


From: Nathan Geffen, 1999. "Cost and Cost-Effectiveness of Mother-to-Child Transmission Prevention of HIV."

3.2.6 Skills development

Apartheid left South Africa with both lower skills levels than most economies at a similar level of productivity, and with highly inequitable access to qualifications based on race. Meanwhile, in the past decade, job losses have affected primarily unqualified work. In these circumstances, skills development - through both basic education and the provision of training and qualifications for adults - is crucial for employment creation, economic growth and equity.

Unfortunately, since 1996 the budget has not reflected this priority, as Table 6 indicates. Education is easily the largest vote on the budget, absorbing over a quarter of total spending. But between 1996 and 1999, in real terms it shrank by 3 per cent a year. The 2000 MTBPS projects only 0,8 per cent growth on average in the coming three years – well below the rate of growth in the population.

Since education is a provincial competency, it is heavily affected by provincial constraints and choices. The poorest provinces spend far less per learner – and also have a lower matric rate.

Table 6. Budget and expenditure per learner by province, 2000


expenditure per learner (rand)

% personnel

learner: educator ratio

matric pass rate


2 650





2 852




Eastern Cape

2 870




Northern Province

3 085




Free State

3 399




North West

3 509




Western Cape

3 682





3 711




Northern Cape

3 885




Source: Department of Finance, Intergovernmental Fiscal Review, 2000

The departments with the lowest budgets per learner also show a relatively high level of spending on personnel. This follows from the national strategy that stabilises learner-teacher ratios within a fairly narrow band. Interestingly, the worst-funded departments employ far fewer non-educators than the Western Cape and Gauteng – yet their low budgets still mean that non-personnel spending is squeezed.

Four main areas of under-resourcing have emerged in education, which current fiscal projections seem unable to address.

Historically African schools, particularly in the former homelands, typically never had clerical support, cleaners or security personnel. As a result, principals, teachers and learners end up carrying out these functions. The Statistics South Africa survey of child labour found that almost 10 per cent of South African children end up doing five hours or more maintenance work at school every week. Some 96 per cent of this group are African, and two thirds live in the rural areas. To equalise the ratio of teachers to support staff between provinces would require the creation of 30 000 unskilled and semi-skilled positions in the schools.

A second problem area lies in the lack of mathematics, science, culture and art teachers in historically African secondary schools. As the National General Council of the ANC pointed out in June 2000, these subjects are particularly important for employment and self-employment in the modern economy. The President listed information and communications technology and cultural industries as two of the growth sectors for the economy this year. Yet most historically African secondary schools still lack teachers for these subjects.

Third, historically African schools, particularly in the rural areas, still lack basic infrastructure. The following table indicates the situation according to the 1996 survey of school needs. Given the decline in education budgets, capital spending comes to only around 2 per cent – far to little to meet these backlogs.

Table 7. Infrastructure backlogs in the schools, 1996


Per cent of the national shortfall of:

Per cent of schools lacking:


Percent of all schools



water in walking distance

flush or im-proved pit toilets



Provinces with large homeland areas (a)








Other provinces (b)








a. Eastern Cape, Northern Province, KwaZulu Natal, North West, Free State and Mpumalanga. b. Western Cape, Gauteng and Northern Cape

The Department of Education is conducting a new School Register of Needs, which is scheduled for release this year. It will help in assessing the progress made in addressing infrastructure backlogs.

The final area of under-funding is Early Childhood Development (ECD) and Adult Basic Education and Training (ABET). These types of education are critical for overcoming the backlogs in education left by apartheid. Yet they lack both systematic programmes and funds. Without modification of the highly restrictive fiscal targets established under GEAR, they probably cannot be contemplated. In the words of the Department of Finance about ECD,

"While eventual returns to such an investment will be high, expenditure will have to compete with other mechanisms for improving quality in the schooling system. It will also place added pressure on provincial education budgets." (Department of Finance, 2000 Intergovernmental Fiscal Review, p. 35)

To deal with the budget shortfall, the education system has permitted schools to levy fees. This system contributes heavily to inequalities. On the one hand, schools in richer areas can charge more, letting them hire more staff and improve their equipment. On the other, poor children cannot afford the fees to these schools. Schools are supposed to take ability to pay into account – but that creates an incentive to avoid admitting poor children at all.

The skills development system established by the Skills Development Act was designed to extend training for adults. That is an especially important because apartheid deprived many adults of competencies and/or qualifications. The new system aims to provide qualifications for adults at all levels of training; to ensure that competencies are planned to meet sectoral needs, rather than being specific to one employer; and to ensure adequate funding of training through a 1-per-cent payroll levy.

When the skills development strategy was under debate, labour argued that the 1-per-cent payroll levy was low. In industrialised and rapidly industrialising countries, studies suggest that spending on training comes to around 4 per cent of payroll. The funding problems may be aggravated by the failure of many businesses to register to pay the levy. Furthermore, government departments only need to budget 1 per cent of payroll for training. There are no consistent mechanisms in place to ensure that they actually spend the funds on training.

The People’s Budget calls for spending on education, like the other components of the social wage, to increase at least 2 per cent above inflation.

3.2.7 Repositioning Infrastructure Spending

Infrastructure expansion remains key to a growth and development path in South Africa. The potential of infrastructure includes:

  • Creating Assets for the Poor: A key lever for eradicating poverty and reducing income inequality is through the provision of viable and sustainable assets for the poor

  • Lower transaction costs: Infrastructure lowers transaction costs by facilitating flows of information and goods, and interactions between markets

  • Creating Economic Linkages: Infrastructure investment creates the potential for economic linkages. In particular, the ability to move goods makes investment viable

  • Concentration of Economic Activity: The provision of infrastructure concentrates economic activity spatially, thus supporting backward and forward linkages

  • Responding to change: Depending on the quality of infrastructure delivered, economies undergoing restructuring are able to respond to shocks, competitive pressures and value-added production

  • Improving productive capacities: Access to infrastructure services could improve the capacities for producing goods and services in communities

  • Creating wealth: Irrigation systems, transport routes and other infrastructure outcomes hold the potential for creating viable assets and markets

  • Jobs: Infrastructure expansion creates jobs both during construction and for maintenance, and by supporting other economic activities in the long run.

  • Boosting demand: Infrastructure expansion also boosts demand in the economy, thus supporting forward linkages. (Hassen E.K. and Horton C, 2000)

Consequently, there is wide agreement that infrastructure expansion is a correct path for government. Yet the budgets for key infrastructure programmes have been declining in real terms, as indicated in Table 4 above. The MTBPS projects an increase for infrastructure in the next three years – but it is still only enough to catch up with 1996 levels of spending around 2006.

To ensure that infrastructure meets developmental challenges requires repositioning of infrastructure programmes in several ways.

First, the current infrastructure programme is not aligned to a coherent trade and industrial strategy. The delivery of infrastructure should go beyond meeting basic needs alone to providing a basis for household economic activity. For example, delivering a communal standpipe or 8 amps of electricity does meet the requirements of a basic need.7 But it is too little to release productive capacities in poor communities by supporting business activity.

Second, the quality of housing spending requires particular focus. As a result of both a real decline in subsidies and poor planning systems, existing housing programmes perpetuate settlement far from urban centres and employment. That has, in turn been associated with:

  • Higher transport costs for workers,
  • Lower population densities, which makes it harder to provide government services and support economic activity,
  • Falling building standards in extension projects, as private sector contractors cut corners, and
  • A failure to locate and plan new settlements in ways that can crowd in other government investments.

Third, rapid implementation of the ANC’s commitment to provide a basic level of service free of cost will both boost the social wage and provide for redistribution. Currently, tariff structures for infrastructure have not supported cross-subsidies from the rich to the poor. As a result, poor communities have faced high bills and shut-offs for non-payment.

Implementing the commitment to free basic services requires important decisions about both budgets and administrative practices. Critical choices relate to the method and level of subsidisation. Above all, it is critical that municipalities in poor regions get national subsidies, since they cannot afford adequate services purely through internal cross-subsidisation.

Fourth, the maintenance of infrastructure is as important as new capital budgets. Table 8 below indicates the costs for rehabilitation of roads over time. The graph indicates that the costs of maintaining roads increase exponentially over time.

Table 8. Costs of road rehabilitation

Road Quality

Time Period for Repair (Approximates)

Cost per km


2-3 years

0,1 mil/km


4 years

1,8 mil/km

Source: Department of Finance, 2000

The most recent calculations of maintenance costs for infrastructure delivery – which should be updated by government - point to a large gap between projected spending and required levels of spending to maintain infrastructure. (See Table 9) The question of investment must further be posed within the context of the call for accelerated service delivery. The financial commitment required to eradicate maintenance and rehabilitation backlogs in the provision of infrastructure has been estimated between R 47 billion and R53 billion in the next five to ten years. Under current medium-term allocations, a shortfall of over R10 billion will emerge.

Table 9. Backlogs in maintenance of public infrastructure


Total backlog and estimated years of spending

Annual shortfall (R billions)

Public Works

R 8,8 billion over five years

R 1.5 billion


R 13 billion for ten years

R 2.4 billion


R 14-20 billion for nine years

R 1.6 billion

Municipal and rural infrastructure

R 45-77 billion for five years

R 10.billion


R 38 billion for ten years

R 5.1 billion


R 47-53 billion

R 10.6 billion

Source: Department of Finance: 1998

Finally, the problems identified with existing infrastructure programmes will be aggravated by strategies pushed by the national Department of Finance to encourage greater private investment in and competition around infrastructure provision. These strategies seem likely to limit the extension of quality infrastructure to the poor.

To compensate for the reduction in budgeted spending on infrastructure, the Department of Finance has demanded that local government and provincial departments shift to so-called public-private partnerships. Under this strategy, the private partner is expected to invest in infrastructure in return for a share in the revenues and/or subsidies. But the strategy is not viable for serving poor communities, where people simply cannot afford to pay the cost of infrastructure. Even if the government provides subsidies to the private partner, experience suggests that it frequently lacks the capacity needed to enforce agreements on service standards.

Fiscal concerns have also led the Department of Finance to push for partial privatisation of the state-owned enterprises that provide infrastructure, especially Telkom and Eskom. As part of this process, these parastatals will be made to compete with private providers. The argument is that this will compel them to become more efficient.

That belief ignores the deep-seated imperfections found in South African markets, and particularly the massive income inequalities. In these circumstances, demanding that state-owned enterprises compete will effectively compel them to focus on the profitable part of the market – supplying the rich and businesses – while scrimping on services to the poor. This shortcoming has been admitted by the Department of Public Enterprise’s own policy documents.

The Departments of Finance and Public Enterprises argue that they can overcome this problem through subsidies to provide for poor consumers. They support this approach on the grounds that it is more transparent than cross-subsidisation of poor customers by rich customers within parastatals.

Given fiscal constraints, however, this strategy seems likely to lead, ultimately, to reduced funding for infrastructure for the poor. Thus, the Department of Finance supported a law to tax Eskom, arguing that instead of using internal cross subsidies to fund electrification, it would provide an equivalent subsidy. In the event, in the coming budget it reportedly plans to give Eskom 40 per cent less than the corporation spent on electrification in 1999/2000.

The People’s Budget argues that expenditure on infrastructure should be increased substantially, as part of the overall social wage. Existing programmes should be reviewed to ensure they contribute as much as possible to both social and economic development.

Furthermore, the emphasis on public-private partnerships at local level, and the proposals for restructuring Telkom and Eskom should be revised to ensure that they will permit government to continue to meet its responsibilities, maintain services for historically underserved communities, provide sustainable security for workers, and fit into a coherent trade and industrial strategy.

  1. Chapter 4: Translating Vision into Reality

This section of our proposals deals with

  • methods to finance our proposals and
  • a framework to improve the capacities of the public service to spend more effectively and efficaciously.

4.1 A developmental fiscal package

The developmental fiscal package that we propose consists of proposals to expand government’s resource base while making the overall tax burden more progressive. It also includes measures to redirect the funds available for spending on the social wage and economic infrastructure. The proposals are:

  1. To increase the level of tax and deficits relative to the GDP,
  2. To shift to a multi-rate VAT, with higher rates on luxuries and more zero-rated necessities,
  3. To restructure the Government Employee Pension Fund (GEPF), and
  4. To review the military procurement programme with a view to reducing it as far as possible.

4.1.1 Increasing the level of tax and deficits relative to GDP

The proposals made here, and in particular the increase in the social wage, requires an expansion in government spending. In the short run, that implies a moderate rise in the ratio of taxes and deficits beyond the targets provided in the MTBPS. In the longer run, as the economy begins to grow more robustly, the deficit and tax ratios will again decline.

EPRI has modelled the implications of this proposal for the People’s Budget. We here present three possible scenarios. All of them increase both taxes and government borrowing, as reflected by the increase in government revenue and the deficit as a percentage of the GDP. The resources raised through higher borrowing and taxes are channelled into productive social spending.

The results of these policies are summarised in the table below, and compared to the projections in the current MTBPS.

Table 10. Scenarios on deficit and tax targets






MTBPS projections

government revenue (% of GDP)





government expenditure (% of GDP)





fiscal deficit (% of GDP)





public debt (% of GDP)





real GDP growth rate






1. gradual expansion

government revenue (% of GDP)





government expenditure (% of GDP)





fiscal deficit (% of GDP)





public debt (% of GDP)





real GDP growth rate





additional social investment (R billions)


R14 bns

R27 bns

R44 bns

2. moderate expansion

Government revenue (% of GDP)





Government expenditure (% of GDP)





fiscal deficit (% of GDP)





public debt (% of GDP)





real GDP growth rate





additional social investment (R billions)


R24 bns

R32 bns

R45 bns

3. rapid expansion

Government revenue (% of GDP)





Government expenditure (% of GDP)





fiscal deficit (% of GDP)





public debt (% of GDP)





real GDP growth rate





additional social investment (R billions)


R33 bns

R 42 bns

R56 bns

*modelled by EPRI at the request of the People’s Budget.

In all the proposed scenarios, the real growth rate increases substantially as a result of

  • Improvements in the social wage, which stimulate domestic demand and increases capital and labour productivity, and
  • A targeted growth strategy that encourages investment and employment creation.

The longer-term results reflect more fully the positive effects of social investment on expanded economic capacity. The real growth rate rises to 5% over the three-year horizon instead of fluctuating at lower levels. In spite of the higher levels of borrowing, the public debt (measured relative to national income) falls more rapidly than in the baseline scenario, since the growth rate of the economy is significantly higher. For instance, even in the least ambitious scenario, in 2002/03 public debt amounts to only 42.7% of GDP, rather than 43% in the baseline scenario, while resources available to finance social investment rise, from R13.7 billion in 2000/01 to R44.4 billion in 2002/03.

4.1.2 Multiple VAT rates

Value-added tax, or VAT, is a highly regressive form of taxation, with the poor paying a higher percentage of their income than the rich. Table 11 indicates the VAT burden on households by income level. It shows that households earning R1500 a month pay 10 per cent of their income on VAT, compared to 7 per cent for those earning over R10 000 a month.

Table 11. VAT burden on households, by income level

Annual House-hold Income

Total VAT Paid in Rands

VAT Paid as a % of total tax paid

VAT Paid as a % of Annual Income

R 18 000

1 799



R 30 000

2 910



R 75 000

6 141



R 140 000

10 241



Source: Department of Finance

In the past three years, tax cuts have concentrated on personal income tax and corporate taxes. Since only those earning over around R2000 a month pay income tax, these reductions do not benefit the poor. Ironically, even the reduction of personal income tax has been regressive, with much greater relief for the high-income group.

Private capital and some opposition parties have been calling for a VAT increase, both to expand the revenue base of government and to fund a basic income grant. The People’s Budget rejects the idea of funding programmes for the very poor by taxing the poor. That would not assist in taking our country out of the poverty trap, or really address the fundamental inequalities that characterise our economy.

In order to deal with the regressive nature of VAT, many countries exempt necessities and impose higher rates on luxury goods. Although many theorists argue for a single, uniform rate, only 18 countries have actually adopted this approach. (COSATU, 1999) Thus, Belize, Canada, Ireland, Jamaica, Kenya, Poland, Romania, Trinidad and Tobago and the United Kingdom zero-rate basic goods, while a further 76 countries have special low rates for basic foodstuffs. Many countries have two or more VAT rates (see Appendix 1).

The People’s Budget proposes an extension of zero rating to additional basic necessities beyond those already zero-rated by government,8 and an increase in VAT on luxury goods.

The actual choice of additional goods to be zero-rated should be informed by current expenditure patterns, with a prioritisation of goods that can substantially contribute to improved social welfare of the poor. To minimise the impact on revenue, commodities with a high value would be a relatively low priority for zero-rating. The health impact of goods – both positive and negative – would also influence the choice of items for further VAT zero-rating. For example, although lower income groups spend a greater proportion of their income on tobacco than do the higher income groups, negative health effects would rule out zero-rating it.

The weighted equity gain ratios cited by the Katz Commission9 suggests the following goods as prime candidates for zero-rating: paraffin, bread flour, white sugar, matches, candles, coal, coal stoves, and white bread. Paraffin is the single clearest candidate. According to calculations of the Katz Commission, the gain per annum to the poorest category of households would be over 15 times the gain to the wealthiest category of households, in absolute terms. Over the counter drugs, generic drugs, and items of the government’s essential drugs list should also be considered for zero-rating, as should school clothes and other education needs.

The increase in the number of goods to be zero-rated would obviously have revenue-loss implications. For this reason, the People’s Budget proposes the introduction of higher VAT rates on luxury goods. This measure would compensate for the decline in revenues.

Certain goods tend to be mainly or almost exclusively consumed by the upper income brackets. These goods are typically relatively import and capital intensive, so that suppressing their demand through higher taxes would have relatively little impact on the domestic economy. It is proposed that two lists of goods be agreed upon: those that prima facie qualify for a luxury tax; and those that qualify for a luxury tax above a certain price threshold.

The first list would include, for instance, video cameras and recorders, decoders, satellite dishes, furs, air conditioners, caravans, yachts and microwaves. The second list would include cars, motorcycles, fridges, freezers, stoves, radios, TVs, watches, jewellery, sunglasses, cosmetics, and furniture. Such a system is already in place in other countries. For example in the United States a car costing over $30 000 is subject to additional tax on the difference between its price and this threshold.

The efficiency of tax collection should also be a consideration. It would be undesirable to have a system with excessive administrative costs. From this point of view it may be preferable to have fewer items subject to luxury tax at a higher rate, rather than a high number of items at a lower rate. We also recommend that below a minimum cost, no luxury taxes are levied.

Combining an extension in VAT zero-rating with increased VAT on luxury goods should mitigate the regressive burden of VAT without affecting overall government revenues.

4.1.3 Restructuring the Government Employees Pension Fund

A third leg of the developmental fiscal package is to restructure the Government Employee Pension Fund (GEPF). The Presidential Jobs Summit agreed on the importance of reviewing the funding of public servants’ employees as a way to release resources to increase social spending. Most public service unions have supported this review.

The GEPF is a defined-benefit fund – that is, employees’ benefits are guaranteed. If the GEPF’s investments do not return enough to cover retirement costs, they must be met out of the budget. The evidence is that it would be cheaper to reduce the funding level of the GEPF and turn the funds to developmental uses, meeting the cost of pensions where necessary from the budget.

The GEPF is currently funded at between 80 and 96 per cent, depending on the investment strategy. The level of funding has risen extraordinarily quickly since 1996, when it was at 60 per cent. This indicates that the current rate of employer contribution – now at 15 per cent of salaries, or about R9 billion a year – is too high.

In the light of this situation, a modest reduction in the level of funding could release substantial resources. The People’s Budget proposes:

  1. The employer contribution could be reduced to 12 per cent of salaries. That will save a total of R1, 5 billion a year.
  2. The funding level could be reduced by 2 per cent, or R3 billion a year, for the next three years, and the funds utilised for once-off investments in infrastructure and housing.

The total increase in funds available for other purposes would thus come to R4, 5 billion a year for the next three years. A more substantial reduction in the level of funding, establishing a fully Pay-As-You-Go system, would release even more resources.

4.1.4 Arms procurement

The arms procurement programme, currently estimated at R43 billion over the next ten to twenty years, effectively shifts spending from the social wage to defence. Since the contracts are mostly denominated in dollars, if devaluation is more rapid than expected the impact will be even greater.

According to the MTBPS, defence spending rose from 6,7 per cent of total spending in 1999/2000 to 7,5 per cent in 2000/1, and will stabilise at 7,7 per cent for the next three years. Defence will grow by 8,5 per cent a year over the period, compared to 7,6 per cent for expenditure as a whole, and only 6,6 per cent for the main social services and the criminal justice system.

It is instructive to compare these expenditures with the MTBPS’s official priorities: infrastructure, the criminal justice system, the social services, provision for HIV/AIDS and employment creation. Defence does not appear on the list – yet its budget will grow faster. The magnitude of the defence spending appears when we look at total expenditure on programmes for "poverty relief and job creation," which the 2000 Budget Review gives as around R5 billion between 1999/2000 and 2002/3.

An unexpected deterioration in the exchange rate would have even more severe implications for the social wage. Some 85 per cent of the package is denominated in foreign currencies. Given the expected slow real growth in the budget, rapid depreciation would squeeze other functions even more than anticipated.

The risk of depreciation is underlined by some of the cost estimates produced in 1998, when the package was first explored. At that time, some analyses came up with exchange-rate estimates based on a long-term depreciation of 5 per cent a year – leading to a rate of R7, 76 to the dollar in 2006. (Cilliers 1998) In the event, depreciation is the main reason that cost estimates soared from R30 billion in 1999 rand to R43 billion in current rand.

Contractual obligations make it difficult to reduce spending on arms procurement in the short run. The People’s Budget proposes, however, that any further procurement be subject to the strictest possible scrutiny. Government should immediately inform vendors that it will not exercise current options to procure additional armaments under existing contracts. Furthermore, if any evidence of corruption surfaces, South Africa should withdraw from the existing contracts.

4.2 Building Public Service Capacity

Improving government’s ability to intervene decisively to support the proposed development strategy will require significant improvements in its capacity and capabilities. We here focus on ways to improve government spending.

Reports on under-spending in government departments have created the impression that government does not have the capacity to spend its resources. In the event, under-spending came to only about 2 per cent of expenditure by national departments in 1999/2000. More recent figures and data on provinces are not available.

Furthermore, as Table 12 shows, almost 80 per cent of under-spending came from just four departments – Welfare, Health, Trade and Industry and Public Works. In every one of these cases, the programmes that displayed under-spending aimed at providing funds to small businesses or NGOs. In contrast, the big social services had no problem spending their entire budget through their normal delivery systems.

Table 12. Expenditure and under-spending by national departments, 1999/2000

National Departments

adjustments estimate

projected outturn


% of budget

% of national de-partments’ underspend













Water Affairs & Forestry




































Trade & Industry






Public Works






National Votes - Total






The rollovers came primarily from:

  • Welfare – social welfare projects provided by NGOs
  • Health – payments for the nutritional support scheme, which shifted the basis for reimbursement in this year, and other conditional grants to provinces
  • Trade and Industry – incentives to business for investment and exports
  • Public Works – the Community Based Public Works Programme.

Developing capacities in the public service will improve both the efficiency of the programmes that are over-spending, and remove obstacles to areas of under-spending. This section of the report identifies areas of problems in spending money effectively, before moving on to suggest possible solutions.

Three related problems exist. These are

  • an inability to spend budgets,
  • the sustainability of projects, and
  • the developmental outcome of expenditure.

We here address the first two issues, since there is still inadequate evidence available to consider the final question.

4.2.1 Inability to Spend

Various factors inhibit the ability of departments to spend their budgets effectively. They include systemic weaknesses in conditional grants, lack of appropriate management skills, excessive reliance on private partners, the procurement system, and the broader effects of fiscal restraint.

Reliance on private partners: A central reason for the inability to spend resources is the increase dependence on private actors to deliver services. Part of the problem lies in lack of experience and systems to manage these relationships. But in many cases, even better government capacity will not make reliance on private partners work. First, as the arms deal indicates, questions on the probity of politicians will be continuously raised, reducing confidence in the public service and parliament. Second, the management of contracts is a specialist skill. Even with a strong skills development strategy, it will be at least several years before a skilled corps of contract managers will emerge. Third, the South African private sector often lacks capacity to meet contract requirements.

Systemic weakness in conditional grants: The current system of conditional grants cannot support rapid and effective spending. The problem is not the conditional grant system itself, but rather its operation. The procedures used for conditional grants are extremely complex and long. As a result, although the budget is announced in February, the disbursements for conditional grants often occur only between April and June. That shortens the period departments have to use the money.

Management skills: The public service currently lacks a skilled corps of project managers to design, implement and monitor infrastructure extension projects. This results in the absence of forward planning (i.e. planning before the actual allocations are made), sluggish implementation systems and very often project breakdown.

Multi-year projects: Another dimension to the problem is that projects often run over several years. Whilst the Medium-Term Expenditure Framework provides a framework for multi-year planning and budgeting, it is becoming clear that departments are often not aligning projects to budgets. Government officials in the infrastructure departments argue that the budget is still too rigid to support project planning.

Design of project systems: Overly complex arrangements that impose enormous delays on spending are typically justified by the need to prevent corruption and misuse of funds. But all too often, they prevent any spending at all. The Poverty Relief programme illustrates how excessively complex project design inhibits effective spending. In this case, three sets of contracts were needed before funds could be disbursed - between the Departments of Welfare and Finance, between Welfare and the project managers, and between the project managers and the recipient communities.

Procurement systems: A similar logic has led to procurement systems that impose great delays on spending. The government tender process may delay major expenditures by three to six months, and sometimes by years. In 1998, the Presidential Review Commission had to terminate all its work and start it over because the Tender Board argued it had not gone through adequate tender processes in hiring consultants. Usually, just getting a tender announcement in the Gazette takes a few weeks. Similarly, it takes months to fill a management position in the public service, making it difficult to start new programmes quickly.

Fiscal restraint: In itself, the heavy fiscal restraint applied to public service institutions has become a barrier to spending. On the one hand, departments are used to the idea that any saving is good, even at the cost of overall delivery. On the other, budget cuts have led to employment freezes and often untargeted cuts in employment. This has reduced capacity, without improvements in efficiency. As a consequence, the capacity of the public service is further reduced. Ironically, the very act of under-spending leads to further budget cuts. Thus this process can become something of a self-fulfilling prophecy.

4.2.2 Sustainability of Projects and Services

A key problem in the spending of many government departments is the question of sustainability. Improving the quality of spending is a goal that all South Africans should share. Not only does it ensure proper returns on our taxes, but it also means that the national project of reconstruction and development will deliver sustainable results.

Our concerns at this point relate to the projects and services that government provides. Key aspects include the following.

Separation of capital and operating budgets. Government currently delivers a great deal of services and has a good record of extending services. However, once infrastructure projects are built, national government often requires local governments or communities to fund the operating and maintenance of projects. Most government departments use this model of separating capital and operating costs. The result of this separation has resulted in insufficient funds being allocated for project operations and maintenance. The result is that projects have broken down. While many government officials indicate that their responsibility stops at capital aspects of the projects, the failure of projects constitutes a waste of limited resources.

Standards of delivery. One result of fiscal restraint is that standards for delivery have been set at low and unsustainable levels. These issues are discussed in the section on infrastructure.

Externalities. Good government spending must catalyse positive externalities. At present, however, there seems to be very little attention paid to ensuring that different programmes and strategies reinforce each other. In particular, economic and social programmes do not have the synergy needed to initiate a virtuous development cycle.

Service packages in institutions. With the transition to democracy, some deterioration in services was unavoidable with the shift to new public policies – for instance, free medical care for pregnant women and children under six and new curricula in the schools. But the attempt to expand and transform services in the face of falling budgets after 1996 has meant the level and quality has often left much to be desired.

4.3 Towards Solutions

A four-stage model of the assessing the success or otherwise of spending is proposed below:

Box 3. A Framework for Assessing the impact of public spending

Policy Stage

Key Issues

Central Policy Questions

Design and Systems

Appropriate Perspective

Is the policy based on acceptable value system or ideology?


Consensus and Agreement

Is the policy broadly accepted within society?


Budget and Human Resources

Is the design consistent with the capacity of the state?


Administrative Simplicity and Accountability

Is the system designed to provide a smooth implementation process?

Does the system protect against corruption?


Cost and Benefit of Policy

Will the benefit of the policy justify the cost?



Has the policy been implemented? Is it a feasible policy for implementation?


Goals and Objectives

Has the policy reached its goals?


Adequacy and Sufficiency

Will the policy lead to solutions to the problems it hoped to address, or will it create new ones?



Has the policy factored for recurrent costs in maintenance and operation?



Will the policy create new problems, in trying to solve another problem?

Sources: Hassen E.K. (2000) and Smith T.(#)

This approach should help civil society, parliamentarians and other actors in the budget process to ask the right (and often difficult) questions when interrogating the budget. In the remainder of this section, we use it to show how government spending could be improved.

Revamping conditional grants. When the Minister of Finance announced the MTBPS for 2001, he should spell out improvements in the system of conditional grants. These proposals should include a review of the Treasury Regulations to make them more support of this kind of dedicated grant.

Procurement: Tender procedures for government departments are extraordinarily cumbersome. In many critical areas, however, they do not necessarily ensure that the best bidder wins. The main concerns are:

  1. The process is extraordinarily time-consuming.
  2. The regulations are designed for procurement of goods, and do not serve well in acquiring services, especially policy advice and research – which may be critical for the overall success of policies. In particular, where it is difficult to specify clear quality criteria for outputs, the process tends to favour low-cost applicants over those who can actually do a quality job.
  3. The tender process does not sufficiently permit departments to take issues of representivity and the strategy of supporting small and micro enterprise into account.

Feasibility of implementation: All proposals – including those from the People’s Budget – should pass the test of being implementable. Unfortunately, the current process of Integrated Implementation Plans will not deliver the level of information and transparency required for civil society to assess the feasibility of programmes, in terms of implementation.

Benefit and Cost: Currently, the cost of programmes is the paramount criteria for selecting projects. A shift in the mindset of government is required, away from saying things are too expensive, without looking at (a) the benefit to society and (b) methods to raise resources.

Information for assessment: To date, government has generally not released sufficient information to assess the achievements beyond simple descriptive statistics. Even where studies exist, they often remain unpublished.

Sustainability: The key proposals here are for a careful assessment of the sustainability of government programmes. Here again, additional research is required in order to develop firm proposals by the People’s Budget. This will be an area on which we develop clear research proposals.


Cilliers, Jackie. 1998. Defence Acquisitions - Unpacking The Package Deals. ISS. Johannesburg.

COSATU. 1999. Proposals to NEDLAC Public Finance and Monetary Chamber on Increasing the Progressivity of Elements of the Tax System (www.cosatu.org.za)

  • 2000. Response to the 2000/2001 Budget (www.cosatu.org.za)
  • 2000. Submission to the Social Security Committee (www.cosatu.org.za)

Department of Finance. 1996. Growth, Employment and Redistribution: A Macro-economic strategy. Pretoria

  • 1998. Medium Term Review Report on Infrastructure
  • 2000. Budget Review (2000) Pretoria
  • 2000. National Expenditure Survey. Pretoria

Geffen N. 2000. Cost and Cost-Effectiveness of Mother-to-Child Prevention of HIV (emailed copy)

Hassen E.K. 2000. When More Means Less: Low-Income and Housing and Macro-Economic Policy. Isandla Institute and WITS Public and Development Management Project on Housing

  • 2001. Lessons in Acceleration: Assessing Service Delivery in South Africa (Draft) National Labour and Economic Development Institute: Johannesburg

Hensher, M., 2000, Confidential Briefing: The costs and effectiveness of using NVP or AZT for the prevention of mother-to-child transmission of HIV – current best estimates for South Africa. Health Financing & Economics: (confidential document, 19 April 2000).

Haarman C. and Haarman D. 1998. Towards a comprehensive social security system in South Africa. Working Paper prepared for the Congress of South African Trade Unions

Haarmann C. 2000. Social Assistance in South Africa: Its potential impact on poverty. Report submitted in partial completion of a doctorate

Hensher, M. 2000. The costs and effectiveness of using NVP or AZT for the prevention of mother-to-child transmission of HIV – current best estimates for South Africa. Department of Health. Health Financing & Economics. 19 April 2000.

People’s Budget. 2000. Response to the Medium-Term Budget Policy Statement (2000)

Ramos M. 2000. Presentation to the NEDLAC Public Finance and Monetary Chamber at the Metal Box Centre in October 2000.

South African Health Systems Trust. 1996. South African Health Review South African Health Systems Trust: South Africa

United Nations Development Programme. 2000. South Africa: Transformation for Human Development (2000). United Nations Development Programme. Pretoria

South African Reserve Bank, Quarterly Bulletin, , , June 2000

Statistics South Africa. 1999. October Household Survey (1999) Statistics South Africa. Pretoria

Background Papers prepared for the People’s Budget Process.

Hassen E.K. and Horton C. 2000. Infrastructure: Budget Analysis. National Labour and Economic Development Institute: Johannesburg

National Health and Allied Workers Union. 2000. National Health Insurance. National Health and Allied Workers Union: Johannesburg

Sampson M. 2000. Macroeconomics. Economic Policy and Research Institute: Cape Town

Appendix 1.

VAT rates across countries


Lower rates %

Standard rate %

Higher rate %

Western Europe



10; 12




0; 1; 6; 12








0; 5; 6; 12; 17




2.1; 5.5








4; 8




0; 6




0; 4; 10; 16




3; 6; 12




5; 12



Republic of Ireland

0; 3.3; 12.5




0; 2; 3



United Kingdom

0; 8






0; 6; 13




0; 5












0; 1





New Zealand




Africa & Middle East



7; 13







Ivory Coast

0; 11.11
















0; 1







South Africa




South America







7; 12




0; 10



North America





Source: Deloitte and Touche (1997) "VAT Handbook".

In most cases, a zero-rating relates to exports and basic food items, and sometimes to clothing.


The National Steering Committee thanks the following persons and organisations for their research and editing assistance:

  • John Capel SACC
  • Kenneth Creamer
  • Ebrahim-Khalil Hassen NALEDI
  • Claire Horton NALEDI
  • Ishmael Lesufi SANGOCO
  • Neva Seidman Makgetla COSATU
  • Ravi Naidoo NALEDI
  • Tebogo Phadu NEHAWU
  • Mike Sampson EPRI
  • Mark Sweet NEHAWU
  • Doug Tilton SACC
  • Fiona Tregenna COSATU
  • Mawethu Vilana NALEDI

The following persons or organisations made their research findings and publications available:

  • Ned Geffen
  • Treatment Action Campaign
  • Campaign Against Military Spending
  • Professor Geoff Harris

In addition, many, many organisations and individual provided comments and constructive criticisms. A special word of thanks goes to these individuals for creating the intellectual climate in which to develop our ideas.

Much of the research undertaken requires that government officials provide access to information. The openness of many government departments to share information and engage on ideas was commendable. We trust that this relationship will continue in the future.

Finally, we are grateful to NALEDI for providing research co-ordination, logistical support and for assisting in the publication of this document.

A final word of thanks goes to the reader. We look forward to hearing your comments on this document: direct them to naledi@naledi.org.za.

National Steering Committee
People’s Budget

February 2001


ANC African National Congress
COSATU Congress of South African Trade Unions
CPI Consumer Price Inflation
DALY Disability Adjusted Life Year
EPRI Economic Policy Research Institute
GDP Gross Domestic Product
GEAR Growth, Employment and Redistribution: A Macro-economic strategy
GEPF Government Employee Pension Fund
MTBPS Medium Term Budget Policy Statement
mtctp Mother-to-Child-Transmission Prevention
NALEDI National Labour and Economic Development Institute
NHI National Health Insurance
OHS October Household Survey
SACC South African Council of Churches
SANGOCO South African Non-Governmental Organisation
VAT Value-added taxes


  1. Statistics South Africa argues that the ILO definition of unemployment requires that it count any income-generating activity.

  2. That is, activities that provide far too little to live on, adding virtually nothing to the economy or household incomes.

  3. That is, industries that produce art or entertainment.

  4. These estimates use the 1996 Census data for the population by age.

  5. CSAE survey.

  6. Child Health Unit and Health Systems Trust in 1996

  7. The standard that constitutes a "basic level of service" is disputed. This debate lies, however, beyond the scope of this paper.

  8. Basic food items that are zero rated are: brown bread, maize meal, samp, mealie rice, dried mealies, dried beans, lentils, pilchards/sardines in tins, milk powder, dairy powder blends, rice, vegetables, fruit, vegetable oil, milk, cultured milk, brown wheaten meal, eggs, edible legumes and pulses of leguminous plants. The number of zero-rated items has varied from time to time. The current list arose with the increase in the VAT rate from 10% to 14% in April 1993.

  9. This ratio indicates the relative importance of a good in the budget of the lowest income bracket as compared to the budget of the highest income bracket.