Final Report Summary - SHIELD (Strategies for health insurance mechanisms to address health system inequities in Ghana, South Africa and Tanzania)
The ultimate aim of the SHIELD project was to critically evaluate existing inequities in health care systems in three African countries, namely in Ghana, South Africa and Tanzania, and the extent to which health insurance mechanisms could address equity challenges. This research was conducted by:
- evaluating the distribution of the current health care financing burden between socio-economic groups, distinguishing between public and private financing mechanisms, and the factors influencing this distribution;
- evaluating the distribution of health care benefits across socio-economic groups and health system-related factors that influence this distribution of benefits;
- identifying and critically evaluating current experience, as well as options for the likely future development of pre-payment mechanisms for funding health care in Ghana, South Africa and Tanzania, and amongst them;
- developing strategies and policy recommendations on pre-payment funding mechanisms that would most appropriately address the identified health system equity challenges
No comprehensive financing incidence analyses existed yet for African countries. In relation to benefit incidence analyses, some work had been undertaken, but this focused solely on quantifying benefits from public subsidies for health services. The SHIELD project greatly added to existing knowledge by not only undertaking the first set of comprehensive financing incidence analyses within African countries, but also through undertaking the first set of benefit incidence analyses that cover publicly funded health services, as well as services funded through other financing mechanisms. In this way, it was the first study that explored in detail income and risk cross-subsidies within the overall health system. SHIELD then went further to undertake a detailed evaluation of the factors contributing to the financing and benefit incidence patterns within each country. This provided an excellent basis for considering alternative health care financing mechanisms that may address those factors contributing to inequitable financing and benefit incidence.
The three countries turned out to be very different in terms of their level of economic development and the current structure of their health systems. A key issue of interest related to alternative health insurance mechanisms in the three countries. Specifically, Ghana is particularly well known for its wide range of community-based Mutual health organisations (MHOs). Tanzania also has growing experience of community-based pre-payment schemes and is known for its Community health fund (CHF) schemes, the first of which was introduced in 1996. In contrast, South Africa has no experience of community-based pre-payment schemes. Instead, it had substantial private voluntary insurance covering middle and high-income formal sector workers. Such insurance schemes were almost non-existing in Ghana and Tanzania. In addition, all three countries were either planning or implementing some form of mandatory health insurance. South Africa progressed the least, but was at that time in the process of agreeing the policy framework for mandatory insurance. Tanzania introduced a social health insurance for civil servants and then introduced insurance cover for private formal sector workers who belong to the National Social Security Fund (NSSF). Ghana took the boldest steps in relation to mandatory insurance by embarking on a national health insurance system within which district-wide community-based schemes are an integral component. Coverage by the NHI in Ghana has increased quite rapidly. South Africa recently announced its intention to introduce a so-called 'NHI', which intends to introduce an integrated and universal health system that will be largely tax funded.
There were some commonalities and some differences in the key health care financing equity challenges identified in the three countries. A matter of concern in all three countries was the level of out-of-pocket payments. Using the international standard measure of catastrophic health care payments being 10 % or more of total household consumption expenditure, over 5 % of households in Ghana and nearly 2 % of households in South Africa and Tanzania incur catastrophic out-of-pocket payments when using health services. When using the alternative threshold of 40 % of non-food consumption expenditure, 2.4 % of households in Ghana and 1.5 % in Tanzania incur catastrophic payments, while very few do in South Africa.
The financing incidence analysis indicated that overall health care financing is progressive in all three countries. In all countries, the poorest 20 % of the population spent a smaller proportion of their household expenditure on contributing to health care funding compared with higher income groups.
Health care funding as percentage of household expenditure across socioeconomic groups in Ghana, South Africa and Tanzania
There were differences between countries in terms of the relative progressivity of different health care funding sources. Tax is an important source of funding in all countries, and the burden of overall tax payments was progressive in each country, largely due to very progressive personal income taxes. However, while Value added tax (VAT) was regressive in South Africa, it was also progressive in Ghana and Tanzania. Fuel levies were found to be regressive in both Ghana and South Africa. While import duties are a very small share of overall tax revenue in South Africa, they are more sizeable in Ghana and Tanzania, where they were found to be progressive. Out-of-pocket payments for health care were found to be regressive in all three countries. Health insurance contributions were found to be progressive in all three countries when these were evaluated across the entire population. However, in Ghana, contributions to the national health insurance by those outside the formal sector were found to be regressive. In addition, in South Africa the distribution of contributions to private health insurance schemes across those who are members of these schemes was regressive. This was largely due to the flat rate contributions in each case. These findings indicated that while overall health care financing was progressive, efforts should be made to reduce the levels of funding from out-of-pocket payments. It also highlighted that in moving towards a greater reliance on pre-payment funding mechanisms careful attention should be paid to the structure of insurance contributions.
Distribution of percentage share of health service benefits and health care needs across socio-economic groups
In relation to the distribution of the burden of health care financing across socio-economic groups, many of the issues are outside the control of the health sector as they relate to tax policy, which is the domain of the Ministry of Finance. The key factors contributing to the regression of out-of-pocket payments and insurance contributions of those in the informal sector (in Ghana and Tanzania) were that flat rates were used (i.e. there was no differentiation of payment amounts according to ability-to-pay) and that policies to exempt the poor from these payments were ineffective, largely due to the inability to identify the poor. Similarly, private voluntary insurance contributions in South Africa imposed a greater burden on lower-income than higher-income scheme members due to charging flat rate contributions.
The consideration of factors influencing benefit incidence patterns highlighted a range of access barriers that affected poorer and rural communities more than their richer and urban-based counterparts. Affordability was a key constraint, not only in terms of the cost of health services themselves but, often more importantly, due to the cost of transport to facilities. The availability of appropriately trained health professionals, equipment and medicines, in addition to the location of health facilities relative to the communities needing health services was also a key constraint. Poor staff attitudes were a major deterrent to using health services when needed.
Finally, country-specific scenarios of possible future health care financing reform were developed. In Ghana, focus was laid on exploring the implications of alternative interventions to include the informal sector in the NHI. While relatively high levels of NHI coverage were achieved (approximately 60 % of the population), it was apparent that it is becoming difficult to enrol more members outside of the formal employment sector. The government of Ghana proposed introducing a 'one-time payment' for those outside the formal employment sector. The implication of different ways of funding NHI membership for the informal sector once they have made their 'one-time payment' was explored.
In Tanzania, the focus was also laid on how to include those outside the formal sector in some form of health insurance scheme. There are moves to integrate the community health funds (which cover those outside the formal employment sector) with the mandatory NHI, which covers civil servants. In Tanzania, the status quo was compared to expanding insurance coverage to the informal sector without subsidised contributions for those with the least ability-to-pay (i.e. population coverage of about 60 %) and with universal coverage, which requires considerable additional health sector investment.
The South African government committed itself to introducing a universal health system that would be largely tax funded, and such that would not require people to pay for health care at the point of service and such that would cover a reasonably comprehensive package of care. Prior to that, it had been proposed to introduce mandatory social health insurance for all those working in the formal sector and to continue to cover the rest of the population from tax funding. These two alternatives were evaluated and compared to the status quo. The modelling found that the universal coverage option would ultimately lead to the lowest levels of health care expenditure. This was largely due to the greater involvement of private health insurance schemes which had relatively high contribution levels in the other two scenarios. However, the universal coverage option would require substantial additional public spending on health care.
- evaluating the distribution of the current health care financing burden between socio-economic groups, distinguishing between public and private financing mechanisms, and the factors influencing this distribution;
- evaluating the distribution of health care benefits across socio-economic groups and health system-related factors that influence this distribution of benefits;
- identifying and critically evaluating current experience, as well as options for the likely future development of pre-payment mechanisms for funding health care in Ghana, South Africa and Tanzania, and amongst them;
- developing strategies and policy recommendations on pre-payment funding mechanisms that would most appropriately address the identified health system equity challenges
No comprehensive financing incidence analyses existed yet for African countries. In relation to benefit incidence analyses, some work had been undertaken, but this focused solely on quantifying benefits from public subsidies for health services. The SHIELD project greatly added to existing knowledge by not only undertaking the first set of comprehensive financing incidence analyses within African countries, but also through undertaking the first set of benefit incidence analyses that cover publicly funded health services, as well as services funded through other financing mechanisms. In this way, it was the first study that explored in detail income and risk cross-subsidies within the overall health system. SHIELD then went further to undertake a detailed evaluation of the factors contributing to the financing and benefit incidence patterns within each country. This provided an excellent basis for considering alternative health care financing mechanisms that may address those factors contributing to inequitable financing and benefit incidence.
The three countries turned out to be very different in terms of their level of economic development and the current structure of their health systems. A key issue of interest related to alternative health insurance mechanisms in the three countries. Specifically, Ghana is particularly well known for its wide range of community-based Mutual health organisations (MHOs). Tanzania also has growing experience of community-based pre-payment schemes and is known for its Community health fund (CHF) schemes, the first of which was introduced in 1996. In contrast, South Africa has no experience of community-based pre-payment schemes. Instead, it had substantial private voluntary insurance covering middle and high-income formal sector workers. Such insurance schemes were almost non-existing in Ghana and Tanzania. In addition, all three countries were either planning or implementing some form of mandatory health insurance. South Africa progressed the least, but was at that time in the process of agreeing the policy framework for mandatory insurance. Tanzania introduced a social health insurance for civil servants and then introduced insurance cover for private formal sector workers who belong to the National Social Security Fund (NSSF). Ghana took the boldest steps in relation to mandatory insurance by embarking on a national health insurance system within which district-wide community-based schemes are an integral component. Coverage by the NHI in Ghana has increased quite rapidly. South Africa recently announced its intention to introduce a so-called 'NHI', which intends to introduce an integrated and universal health system that will be largely tax funded.
There were some commonalities and some differences in the key health care financing equity challenges identified in the three countries. A matter of concern in all three countries was the level of out-of-pocket payments. Using the international standard measure of catastrophic health care payments being 10 % or more of total household consumption expenditure, over 5 % of households in Ghana and nearly 2 % of households in South Africa and Tanzania incur catastrophic out-of-pocket payments when using health services. When using the alternative threshold of 40 % of non-food consumption expenditure, 2.4 % of households in Ghana and 1.5 % in Tanzania incur catastrophic payments, while very few do in South Africa.
The financing incidence analysis indicated that overall health care financing is progressive in all three countries. In all countries, the poorest 20 % of the population spent a smaller proportion of their household expenditure on contributing to health care funding compared with higher income groups.
Health care funding as percentage of household expenditure across socioeconomic groups in Ghana, South Africa and Tanzania
There were differences between countries in terms of the relative progressivity of different health care funding sources. Tax is an important source of funding in all countries, and the burden of overall tax payments was progressive in each country, largely due to very progressive personal income taxes. However, while Value added tax (VAT) was regressive in South Africa, it was also progressive in Ghana and Tanzania. Fuel levies were found to be regressive in both Ghana and South Africa. While import duties are a very small share of overall tax revenue in South Africa, they are more sizeable in Ghana and Tanzania, where they were found to be progressive. Out-of-pocket payments for health care were found to be regressive in all three countries. Health insurance contributions were found to be progressive in all three countries when these were evaluated across the entire population. However, in Ghana, contributions to the national health insurance by those outside the formal sector were found to be regressive. In addition, in South Africa the distribution of contributions to private health insurance schemes across those who are members of these schemes was regressive. This was largely due to the flat rate contributions in each case. These findings indicated that while overall health care financing was progressive, efforts should be made to reduce the levels of funding from out-of-pocket payments. It also highlighted that in moving towards a greater reliance on pre-payment funding mechanisms careful attention should be paid to the structure of insurance contributions.
Distribution of percentage share of health service benefits and health care needs across socio-economic groups
In relation to the distribution of the burden of health care financing across socio-economic groups, many of the issues are outside the control of the health sector as they relate to tax policy, which is the domain of the Ministry of Finance. The key factors contributing to the regression of out-of-pocket payments and insurance contributions of those in the informal sector (in Ghana and Tanzania) were that flat rates were used (i.e. there was no differentiation of payment amounts according to ability-to-pay) and that policies to exempt the poor from these payments were ineffective, largely due to the inability to identify the poor. Similarly, private voluntary insurance contributions in South Africa imposed a greater burden on lower-income than higher-income scheme members due to charging flat rate contributions.
The consideration of factors influencing benefit incidence patterns highlighted a range of access barriers that affected poorer and rural communities more than their richer and urban-based counterparts. Affordability was a key constraint, not only in terms of the cost of health services themselves but, often more importantly, due to the cost of transport to facilities. The availability of appropriately trained health professionals, equipment and medicines, in addition to the location of health facilities relative to the communities needing health services was also a key constraint. Poor staff attitudes were a major deterrent to using health services when needed.
Finally, country-specific scenarios of possible future health care financing reform were developed. In Ghana, focus was laid on exploring the implications of alternative interventions to include the informal sector in the NHI. While relatively high levels of NHI coverage were achieved (approximately 60 % of the population), it was apparent that it is becoming difficult to enrol more members outside of the formal employment sector. The government of Ghana proposed introducing a 'one-time payment' for those outside the formal employment sector. The implication of different ways of funding NHI membership for the informal sector once they have made their 'one-time payment' was explored.
In Tanzania, the focus was also laid on how to include those outside the formal sector in some form of health insurance scheme. There are moves to integrate the community health funds (which cover those outside the formal employment sector) with the mandatory NHI, which covers civil servants. In Tanzania, the status quo was compared to expanding insurance coverage to the informal sector without subsidised contributions for those with the least ability-to-pay (i.e. population coverage of about 60 %) and with universal coverage, which requires considerable additional health sector investment.
The South African government committed itself to introducing a universal health system that would be largely tax funded, and such that would not require people to pay for health care at the point of service and such that would cover a reasonably comprehensive package of care. Prior to that, it had been proposed to introduce mandatory social health insurance for all those working in the formal sector and to continue to cover the rest of the population from tax funding. These two alternatives were evaluated and compared to the status quo. The modelling found that the universal coverage option would ultimately lead to the lowest levels of health care expenditure. This was largely due to the greater involvement of private health insurance schemes which had relatively high contribution levels in the other two scenarios. However, the universal coverage option would require substantial additional public spending on health care.