A Comparative Analysis
of the Financing of
HIV/AIDS Programmes
in Botswana, Lesotho, Mozambique,
South Africa, Swaziland and Zimbabwe
OCTOBER 2003
Prepared for the Social Aspects of HIV/AIDS and
Health Research Programme of the
Human Sciences Research Council
by Dr H. Gayle Martin
Funded by the WK Kellogg FoundationFree download from www.hsrc
p
ublishers.ac.za
Contents
Executive Summary vii
Acknowledgements xi
Abbreviations xii
Introduction 1
Methodology 3
Definition of HIV/AIDS Expenditures 3
Data Collection 4
Limitations and Challenges 5
Botswana 7
Level of Expenditure 7
Functional Classification of HIV/AIDS Expenditures 11
Sources of Financing 11
Financing Mechanisms 11
Lesotho 15

Total Expenditure on HIV/AIDS 49
Conclusion 51
Special Resource Mobilisation Strategies 51
Do Increased Resources mean Increased Inefficiency? 51
Sustainability 51
Appendices 53
Appendix A: Selected Indicators by Country 53
Appendix B: HIV/AIDS Indicators by Country 57
Appendix C: Terms of Reference 58
Bibliography 59Free download from www.hsrc
p
ublishers.ac.za
Tables
Table 1: Total and Public Health Expenditure in Botswana (1990–2000) 8
Table 2: Core Expenditure on HIV/AIDS Programs in Botswana (1999/01–2002/03, in
current US$) 9
Table 3: Sources of Funding for HIV/AIDS programmes in Botswana (2000) 10
Table 4: Expenditure on HIV/AIDS in Botswana (2001/02) 10
Table 5: Functional Classification of Government of Botswana HIV/AIDS Expenditure
(2002/03) by Financing Mechanism 13
Table 6: Global Fund Award to Botswana 13
Table 7: Total and Public Health Expenditure in Lesotho (1990–2000) 15
Table 8: Government of Lesotho funding for HIV/AIDS, Tuberculosis and Malaria in
(2001/02) 16
Table 9: Expenditure on HIV/AIDS in Lesotho (2001/02) 16
Table 10: External Sources of Funding for HIV/AIDS programmes in Lesotho (2000) 17
Table 11: Global Fund Award to Lesotho 19

Table 33: Summary of Expenditure on HIV/AIDS by Country (2000/01, US$) 46
Table 34: Summary of Expenditure on HIV/AIDS by Country
(2000/01, International $) 47
v
©HSRC 2003Free download from www.hsrc
p
ublishers.ac.za
Figures
Figure 1: Total health expenditure (US$) per capita) in the six countries vii
Figure 2: Health expenditure in the six countries as a percentage of government
expenditure (2001/02) viii
Figure 3: Total HIV/AIDS expenditure (US$ millions) ix
Figure 4: Total HIV/AIDS expenditure (US$) ix
Figure 5: Change in Life Expectancy in Botswana (1970–2000) 7
Figure 6: Financial Flows for HIV/AIDS Expenditure in Botswana 12
Figure 7: Financial Flows for HIV/AIDS Expenditure in Lesotho 19
Figure 8: Sources of Health Financing in Mozambique (1997) 22
Figure 9: The Flow of Resources for HIV/AIDS to the Provincial Level in
South Africa 29
Figure 10: Total Health Expenditure (A) as a Percentage of GDP and (B) Per Capita
(US$) for 1990–2000 By Country 44
Figure 11: Health Expenditure as a percentage of government expenditure
by Country 45
Figure 12: Government expenditure on HIV/AIDS per capita and per PLWHA (2001) 47
Figure 13: Expenditure on HIV/AIDS as a percentage of GDP (2001) 48
Figure 14: Expenditure on HIV/AIDS (2001/02, current US$) 48
Figure 15: Share of Government and External Sources of HIV/AIDS Financing 49

eighty per cent of the infected adults in these six countries.
It is within the context of this HIV/AIDS burden that this comparative analysis aims to
assess the readiness and ability of the countries to respond to the HIV/AIDS epidemic.
The key issues that are addressed in this analysis are:
• Is the allocation to health, as a per cent of total government expenditure, sufficient?
• Is enough allocated to deal with HIV/AIDS, given the magnitude of the problem?
Data limitations made it nearly
impossible to evaluate HIV/AIDS
expenditure allocation – in terms of
economic classification (capital and
recurrent) or functional classification
(prevention, care and support, and
treatment). The allocation of HIV/AIDS
funds by activity is therefore, generally,
not addressed in the report.
Another data limitation was the paucity
of information on household (and
business) expenditure on HIV/AIDS.
Estimates from Latin American and
Caribbean countries found that average
annual expenditure by people living
with HIV/AIDS (PLWHA) was US$1,000,
while an assessment in Rwanda reported US$25 per PLWHA. Even at the latter level, it is
clear that significant amounts of household resources are devoted to HIV/AIDS, resulting
in a combination of transient and permanent impacts on household welfare. One
particular outcome is an increase in the number of households falling below the poverty
line. While not addressed in this report, this household-level outcome has several
secondary consequences that also need to be considered – for example, increasing the
demand for government assistance in the form of poverty alleviation.
vii

health services range from a low of US$12 (in the World Development Report 1993) to
US$34 (by the Macroeconomic Commission on Health in 2001). Four of the six countries
have expenditures in excess of these levels, although two countries, Lesotho and
Mozambique, have per capita expenditures of well below US$34, and in the case of
Mozambique, below US$12.
Is the allocation to health, as a per cent
of total government expenditure,
sufficient? Except for South Africa and
Zimbabwe, none of the countries fulfilled
their commitment made in Abuja in April
2001 to allocate 15 per cent of
government expenditure to health.
Botswana comes closest among the
remaining countries, spending ten per
cent of government expenditure on
health. The other countries spend about
half of the 15 per cent target. It should
however be noted that this data is for the
years 2001 and 2002. When viewed
against the background of increasing
allocations to the health sector over time, it is likely that Botswana and Swaziland will
meet the target. However, the constrained macroeconomic environment in Mozambique
and Lesotho suggests less optimism for reaching the targeted 15 per cent.
Aggregate government expenditure on HIV/AIDS in these southern African countries is
nearly US$70 million annually. There is great variation in the level of expenditure on
HIV/AIDS by individual countries. Government expenditure on HIV/AIDS ranges from a
high of US$33 million in South Africa to a low of US$0.8 million in Lesotho. Per capita
expenditure on HIV/AIDS shows similar variation – on the high end is Botswana with
US$30 per capita, which is almost 30 times the level of expenditure in the other countries.
All the other countries fall below US$1.50 per capita. The median per capita HIV/AIDS


Free download from www.hsrc
p
ublishers.ac.za
Executive summary
these countries are financed mainly by
external sources. In Mozambique, Lesotho
and Swaziland more than eighty per cent
of total HIV/AIDS spending is funded by
external sources. The allocations from the
Global Fund to Fight AIDS, Tuberculosis
and Malaria to these countries will add an
additional US$479 million over the total
period of the allocations, and US$192 over
the first two years of each award.
Total spending in these countries
(government- and donor-financed but
excluding household out-of-pocket
spending and the Global Fund allocations)
amounts to approximately US$250 million
for the year 2001, or to US$3 per capita
and US$27 per PLWHA. In the literature,
the reported HIV/AIDS spending per capita
(excluding out-of-pocket spending) for
sub-Saharan Africa is US$0.3 per capita and
US$8 per PLWHA. Regardless of the
measure, total expenditure on HIV/AIDS in
these six countries is higher than the
regional average. Specifically, per capita
HIV/AIDS expenditure is ten times higher

30
25
20
15
10
5
0
$1
$8
$2
$19
External
Government
Figure 4: Total per capita HIV/AIDS expenditure
(US$)Free download from www.hsrc
p
ublishers.ac.za
The financing of HIV/AIDS programmes
The gravity of the HIV/AIDS situation in these six countries calls for prioritisation,
protection and targeting of HIV/AIDS spending. What is the appropriate institutional
funding mechanism for responding to this call? A detailed assessment of the experiences
of, for example, Zimbabwe (with the earmarking of three per cent wage tax for HIV/AIDS
expenditures), Lesotho (with the allocation of two per cent of all sectoral budgets to
HIV/AIDS) and South Africa (with the introduction of a conditional grant for HIV/AIDS),
is required in order to make specific recommendations. However, preliminary evidence
suggests that the experiences of Zimbabwe and South Africa have generally been positive,
although Lesotho has had less success. Some of these experiences are shared in the

This report is the product of contributions from various research teams. I would like to
acknowledge them and their contributions to making this monograph possible. The data
collection was completed because of the joint efforts of research teams in Botswana,
Lesotho, Mozambique, South Africa, Swaziland and Zimbabwe. The team leaders were:
•Professor Sheila Tlou – Botswana
•Dr Ron Cadribo – Lesotho
•Mr Joel Gudo – Mozambique
•Efua Dorkenoo – South Africa
• Rudolph Maziya – Swaziland
•Brian Chandiwana – Zimbabwe
The Departments of Treasury/Finance and the Departments of Health in the six countries
played an important role in the provision of information, without which this report would
not have been possible. We sincerely appreciate their collaboration.
The role of Dr Gayle Martin in analysing the data, synthesizing information, often
augmenting this with insights gleaned from other sources, and then writing it all up, is
much appreciated.
The editorial and production work of HSRC Publishers will not go unnoticed. They
worked under extreme time pressure and managed to get the report completed within the
given time frame.
Finally, the financial contribution of the WK Kellogg Foundation, and the support of
Bishop Malusi Mpumlwana and Mrs Vuyo Mahlati, who offered constant encouragement
and support throughout the project, is highly valued.
The Social Aspects of HIV/AIDS and Health Research Programme of the Human Sciences
Research Council takes responsibility for the content of this report because it was
responsible for conceiving the idea and ensuring that it was successfully carried out and
completed.
Dr Olive Shisana
Executive Director, Social Aspects of HIV/AIDS and Health Research Programme,
Human Sciences Research Council
Acknowledgements

UN United Nations
UNAIDS Joint United Nations Programme on HIV/AIDS
UNDP United Nations Development Programme
UNFPA United Nations Population Fund
UNICEF United Nations Children’s Fund
USAID United States Agency for International Development
VCT Voluntary counseling and testing
WHO World Health Organization
xii
©HSRC 2003
AbbreviationsFree download from www.hsrc
p
ublishers.ac.za
Early in the twenty-first century governments and world leaders made several
unprecedented, high-level commitments to fight HIV/AIDS. One of the commitments
made by all governments during the Special Session of the United Nations General
Assembly on HIV/AIDS in June 2001 in the Declaration of Commitment was to ‘secure
more resources to fight HIV/AIDS increasing annual spending to US$7-10 billion in low-
and middle-income countries’. The Abuja Declaration, made by African leaders in April
2001 stated: ‘We commit ourselves to take all necessary measures to ensure that the needed
resources are made available from all sources, and that they are efficiently and effectively
utilised. We pledge to set a target of allocating at least 15 per cent of our annual budget to
the improvement of the health sector. We undertake to mobilise all the human, material
and financial resources required to provide care and support and quality treatment to our
populations infected with HIV/AIDS, tuberculosis and other related infections.’ This
commitment was endorsed at the UNGASS and world leaders from developed countries
also committed to assist African leaders in their efforts to realise the funding targets set in

ublishers.ac.za
The financing of HIV/AIDS programmes
2
©HSRC 2003Free download from www.hsrc
p
ublishers.ac.za
The purpose of this study was to identify the sources and quantities of the funds available
for health expenditure and, more specifically, for HIV/AIDS programmes in the six
countries. The study sought to identify the different sources of funds for HIV/AIDS
interventions that are available to the Ministry of Finance/National Treasury. These
sources have been dis-aggregated as far as the data would allow. The purpose of the
study was, however, not to scrutinise the allocation criteria but to quantify the allocation
for health and HIV/AIDS.
None of the countries studied have undertaken a ‘National HIV/AIDS Accounts’, as has
been done in other countries such as: Rwanda, Argentina, Brazil, and other Latin
American and Caribbean countries. The data are therefore largely incomplete. For
example, private/household expenditures on HIV/AIDS have been completely omitted,
despite several studies indicating that households of PLWHA contribute substantially to
HIV/AIDS expenditures. For example, a study in twelve Latin American and Caribbean
countries found that average expenditure by PLWHA was US$1,000, ranging from over
US$3,000 (in Uruguay) to US$125 (in Guatemala). An assessment in Rwanda reported
US$25 per PLWHA spent on HIV/AIDS-related expenditures. If the latter amount were
extrapolated to the six countries, the level of aggregate household expenditure on
HIV/AIDS would be over US$200 million, if one only considers expenditure by adults
with HIV/AIDS. This estimate would have to be verified in a purposively sampled
household survey and facility-based (private and public) survey in a multi-country setting.
The high level of out-of-pocket healthcare spending that households already incur

ublishers.ac.za
The financing of HIV/AIDS programmes
The expenditures have been reported in US$ to facilitate cross-country comparison. In
Table 34 a summary table is presented in international dollars (PPP$), allowing for
purchasing power parity. Purchasing power parity takes into account variances in
domestic purchasing power of a given level of expenditure. However, it makes the
assumption that prices in the health sector follow the same structure as the broader
economy, which is not always the case. Variance in domestic and international purchasing
power is especially of concern for non-tradable inputs, for example, domestically
provided labour inputs (which often account for up to 80 per cent in the health sector).
Where goods are purchased from international markets (for example, imported drugs,
imported medical supplies, and international consultants) the conversion to international
dollars does not make sense. This would be the case for donor financing and some of
domestic funding. This would suggest using PPP$ for government funding and US$ for
donor funding. However, this approach would provide internally inconsistent, aggregate
estimates of expenditure, and for this reason the reporting and analysis are done in US$,
while the purchasing power parity conversion is done for mainly for completeness.
Data Collection
The key sources of information were Ministries of Finance/National Treasuries, Ministries
of Planning, Ministries of Health, Tax Offices, National Income Accounts and National
Health Accounts, income/employment and household surveys as well as any special
research studies. An important source of information was the submissions by four of the
countries (Botswana, Lesotho, Mozambique and Swaziland) to the Global Fund. Extensive
use was also made of the reports by multilateral agencies: UNAIDS, UNDP, World Bank
and IMF.
During the data collection the assumption was made that in each country the Ministries of
Finance/National Treasuries are directly responsible for distributing funds to the various
sectors and would therefore be one of the best sources of expenditure information. The
original intention was to divide the sources into:
• public sources (tax and compulsory heath insurance contributions)

• human resource development
•treatment of HIV/AIDS
•HIV/AIDS research
• administration
Funds for HIV/AIDS control could have been allocated to various intermediaries for a
number of uses, as is the practice in National Health Accounts. The intermediaries may
include the following: regional/local governments, non-government organisations (NGOs),
community-based organisations (CBOs), research institutions, and other ministries. Where
available, the intermediaries were reported.
Limitations and Challenges
Discrepancies have been noted in data obtained from different sources. This illustrates
the difficulties encountered in collecting the data. The experience from other studies
assessing HIV/AIDS expenditures showed that questionnaires to government institutions
and HIV/AIDS co-ordinating structures are an inefficient data collection tool. This was
also the experience in this study.
The limitation of data from special studies is that different definitions may be used in the
variables that are reported, or they are not sufficiently dis-aggregated for the purpose at
hand. The lack of information about households’ out-of-pocket expenditure was a further
limitation.
For these reasons, the reported data should be taken as an indication of expenditures on
HIV/AIDS, and an effort should be made to have the various governments and external
partners verify the data captured in this report. Input by donors would be especially
pertinent to minimise the potential for double counting of donor inputs. For example,
UN agencies often implement bilateral donor-funded activities and a simple aggregation
of donor expenditures, as was done in this report, will likely suffer from some degree of
double counting of donor inputs. In future, National HIV/AIDS Accounts would be able
to address the assumptions made in the data just mentioned, as well as data omissions
encountered in this study.
5
©HSRC 2003

due to HIV/AIDS of 0.6 per cent point in 2001 and a loss of 2.7 per cent points by 2015.
Level of Expenditure
Health expenditure
Total health expenditure in Lesotho averaged at about six per cent of GDP for the past
two decades. Public health spending increased from a low of 2.6 per cent in 1990 to
5.2 per cent of GDP in 2000 (Table 7). Per capita spending at US$28 is slightly above the
mean per capita health expenditure of low-income countries ($21).
As alluded to earlier, the HIV/AIDS expenditure for Lesotho needs to be viewed within
the context of macroeconomic concerns, given the lack of government control over
monetary policy. Government will likely place strong emphasis on containment of public
expenditure to address the macroeconomic challenges. While health sector allocations
have not come under threat, it is unlikely that the health budget will show significant real
increases (from government sources) in the medium term.
Government Financed HIV/AIDS Expenditure
In 2001/02 the government of Lesotho spent US$0.8 million on HIV/AIDS programmes
and a further US$0.2 million on tuberculosis and malaria (Table 8). These expenditures
15
©HSRC 2003
Table 7:Total and public health expenditure in Lesotho (1990–2000)
1990 1995 2000
Total health expenditure (percentage of GDP) 6.2 6.3
Public health expenditure (percentage of GDP) 2.6 4.9 5.2
Total health expenditure per capita (current US$) 31 28
Source: World Bank, 2003Free download from www.hsrc
p
ublishers.ac.za
The financing of HIV/AIDS programmes

PLWHA 360,000
Source: Government of Lesotho, 2002Free download from www.hsrc
p
ublishers.ac.za
17
©HSRC 2003
Table 10: External sources of funding for HIV/AIDS programmes in Lesotho (2000)
Name of agency Type of Budget Time period Estimated annual Main programmes
agency (US$) budget (US$)
CHAL (Christian Health NGO 134,191 3 years 44,730 Lesotho Pastoral Education Project, care and support, counselling,
Association of Lesotho) prevention, income generating activities, clinical care, home-based
care, adolescent health
CARE NGO 2,000,000 3 years 666,667 IEC: Sexual Health and Rights Programme (SHARP)
Lesotho National Assoc. for
the Physically Disabled NGO 3,500 3 years 1,167 IEC for People with Physical Disabilities
Lesotho Red Cross Assoc. NGO 78,936 3 years 26,312 Youth against HIV/AIDS
World Vision International NGO 1,500,000 3 years 500,000 Support for orphans and vulnerable children through educational
and Save the Children inputs, material donations and developmental activities
DFID Bilateral 1,500,000 3 years 500,000 Regional AIDS and agriculture
KFW Bilateral 350,000 3 years 116,667 Family planning, condoms, STI drugs
Ireland Aid Bilateral 650,000 3 years 216,667 Bilateral support to HIV/AIDS programme
World Bank Multilateral 2,000,000 3 years 666,667 Multisectoral HIV/AIDS initiatives
World Food Programme Multilateral 5,361,595 3 years 1,787,198 Mitigation of HIV/AIDS
UNFPA Multilateral 1,002,837 3 years 334,279 Population policy review, reproductive health, IEC materials
development, youth centre
WHO Multilateral 858,678 3 years 286,226 Care and support, adolescent friendly health services,
IEC material development

Association, CARE, Lesotho National Association for the Physically Disabled, Lesotho Red
Cross Association, World Vision, Save the Children) (see Table 10).
Financing Mechanisms
Figure 7 demonstrates the flow of funds for HIV/AIDS-related activities. As mentioned, the
two main sources of funding are the Ministry of Finance and Planning and donors. The
Ministry of Finance and Planning allocates resources (as part of the health budget) to the
Ministry of Health and Social Welfare for HIV/AIDS interventions, which are implemented
mainly by the Disease Control Unit in the ministry. These interventions are implemented
in health facilities by NGOs or CBOs. The Ministry of Health and Social Welfare also
receives donor assistance for the implementation of its HIV/AIDS programmes.
Realising the magnitude of the impact of HIV/AIDS, the Ministry of Finance and Planning
has introduced a targeting strategy for HIV/AIDS expenditure that is intended to have
minimal fiscal impact. For the last two financial years (2001/02 and 2002/03), the Ministry
of Finance and Planning has required that each sectoral ministry commit a minimum of
two per cent of their respective budgets to HIV/AIDS-related activities. This has been
largely an unsuccessful tool to target multisectoral resources for HIV/AIDS. Recently, this
directive has become more explicit by dictating the line items for ministries to allocate
resources to HIV/AIDS. The 2002/03 national government budget is approximately
US$437.3 million, and through this initiative, an estimated US$7.8 million is targeted for
HIV/AIDS-related activities (from the budgets of the various ministries). It should,
The financing of HIV/AIDS programmes
18
©HSRC 2003Free download from www.hsrc
p
ublishers.ac.za
however, be noted that the two per cent allocated translates into five per cent of what,
according to LAPCA, is required for implementation of the National AIDS Strategic Plan.

1 & 2 (US$) budget (US$) PLWHA
HIV/AIDS & TB 10,557,000 29,312,000 14.25 81.42
Tuberculosis 2,000,000 5,000,000 2.43 13.89
Total 12,557,000 34,312,000 16.68 95.31
Source: Global Fund, 2003Free download from www.hsrc
p
ublishers.ac.za
The financing of HIV/AIDS programmes
20
©HSRC 2003Free download from www.hsrc
p
ublishers.ac.za
Mozambique
Mozambique is one of the poorest countries of the world. Globally, it has one of the
highest infant mortality rates (130 per 1000 live births) and maternal mortality ratios
(980 per 100,000) (Figure 16 and Figure 17 in Appendix A). After many years of war and
devastation, the economy has recovered significantly, but the floods in 2000 curtailed this
recovery (Figure 21 in Appendix A). Between 1997 and 2000, average per capita income
in the economy increased from US$196 to US$ 224 (at 2000 prices), a real annual growth
rate of about five per cent. Given the population of 17.7 million people this translates
into a per capita GDP of US$220.
Level of Expenditure
Health expenditure
Between 1997 and 2000 total public expenditure on health grew much faster than GDP,


Free download from www.hsrc
p
ublishers.ac.za
The financing of HIV/AIDS programmes
and US$14.58 per PLWHA were spent in 2001. Of the total government expenditure on
HIV/AIDS, US$10.2 million came from the health budget. This is up from US$3.3 million
in 1999.
Externally Financed HIV/AIDS Expenditure
In 2001 Mozambique received external funding for HIV/AIDS of approximately US$73.3
million (Table 14). This is equal to US$4.14 per capita and US$66.64 per PLWHA. External
sources account for 82 per cent of total HIV/AIDS expenditures in Mozambique.
Total HIV/AIDS expenditure
Table 15 shows the total financial resources available for core HIV/AIDS expenditure in
Mozambique. Total expenditure for HIV/AIDS in 2001 was US$89.4 million or 2.40 per
22
©HSRC 2003
Figure 8: Sources of health financing in Mozambique (1997)
Treasury 22%
Donors 52%
Employers 7%
Households 19%
Source: Chao and Kostermans, 2002
Table 13: Government of Mozambique funding for HIV/AIDS, tuberculosis and malaria (2001)
Expenditure type US$ Per capita Per PLWHA Percentage Percentage
US$ US$ of total of GDP
HIV/AIDS 16,036,000 0.9 14.58 68 0.4
Tuberculosis 1,600,000 0.1 0.32 7 0.0
Malaria 6,000,000 0.3 35.29 25 0.2
Total 23,636,000 1.3 10.28 100 0.6