Prioritising adequate public health financing

The commitment made in Abuja by heads of state to 15% of domestic government spending on health was triggered by the demands of HIV, tuberculosis and malaria.17 The demands from rising NCDs, epidemic outbreaks and pandemics, climate-related health shocks and the commitment to UHC suggest, however, that this commitment remains as relevant today as when it was made in 2001. While other sectors contribute to health, the Abuja commitment was regarded as a share that would reflect adequacy and prioritisation for the health sector to play its own role, including within whole of government approaches. Table 1 shows the level of current government health expenditure as a share of total government expenditure.

In 2018, no ESA country had attained the 15% target, although some countries—Lesotho, South Africa and Botswana—were close to it. Seven ESA countries spent half or less than the 15% committed to by heads of state (table 1). A case study from Zambia suggested that the demands of the COVID-19 pandemic on an underfunded health system led to rapid mobilisation of funds above the Abuja commitment of 15% for health sector pandemic control, raising recognition of the need for investment in improved public sector health systems, including for future pandemic preparedness.14

Adequacy of GDP shares and per capita public sector financing for health

Table 1 shows that spending on health exceeded 5% of GDP in 11 ESA countries in 2018, while noting that this combines resources from public, private and household sources. The evidence of public spending levels and OOP expenditure shares shown in table 1 suggest that fewer countries may have achieved this through significant public sector spending. Table 1 indicates the wide variability in per-capita government spending on public sector health services across the ESA region, with very low levels in six ESA countries, and countries with smaller populations and higher overall GDP appearing to fare better.

A number of estimates have been made of the funding needed for health system functioning. We used two WHO estimates and one estimate from country costings of essential health benefit packages to identify costs per capita for the public sector to deliver its health system obligations.

In 2012, WHO estimated that the minimum spending per person per year needed to provide basic, life-saving services was US$44.19 Adjusted for inflation, this translates to US$48/capita in 2018. As table 1 shows, 10 ESA countries spent below this 2018 level, with only countries with small populations and high GDPs spending above it. The average per capita funding gap for those spending below the basic threshold was US$28, ranging from US$9 in Zimbabwe to US$45 in the DRC.

In 2018, selected ESA governments calculated the costs for the public sector to deliver an essential health benefit.21 For the three countries that assessed total costs shown in table 2, the 2018 per capita cost ranged from US$52 to US$560, with the latter figure in Eswatini reflecting its small population as an outlier.

For the majority of the countries in the region, it is reasonable to take the common US$52 found for the two other countries, one in east Africa and one in southern Africa, as a proxy. As shown in table 1, 10 ESA countries spent below this level, with an average per capita funding gap of US$32, ranging from US$13 in Zimbabwe to US$49 in the DRC.

These two prior estimates shown in table 2 are a conservative minimum. The 2000 World Health Report estimated that US$60/capita was needed for a comprehensive health system annually, including a minimal set of interventions and the infrastructure to deliver them. In 2001, this estimate was revised to US$80/capita per year.20 Adjusting for inflation, this translates to US$114/capita in 2018. As table 1 shows, 12 ESA countries with a combined population of 374.2 million24 spent below this level in 2018, with an average per capita funding gap in these countries of US$84, ranging from US$25 in Eswatini to US$111 in DRC.

Hence while five ESA countries—Namibia, Mauritius, South Africa, Botswana and Seychelles—funded their public sectors above all of these recommended per capita expenditure levels, the public sector per capita financing gap for the others ranged from an average per capita level of US$28 for the most conservative estimate of system needs, to US$84 for a more comprehensive system. For these 12 ESA countries, this implied a total shortfall in public financing that ranged from US$10.5 billion to US$31.4 billion. This represents a significant shortfall in public financing to deliver the minimum health service package, and even more so for a comprehensive health system.

The contribution and implications of private and OOP health expenditures

Private financing comes primarily from OOP payments and voluntary private health insurance.5 23 Domestic private expenditure as a share of current expenditure varies widely across ESA countries, from 16% in Mozambique, Malawi, Botswana and Zambia, to over 50% in the DRC, Zimbabwe, Angola and Mauritius (table 1). The share of private expenditure rises not only when private financing rises, but also when public financing falls.

While this may point to a potential contribution from private financing to meet the public sector health financing gap noted above, a rising share of private financing coupled with falling public financing is also documented in the region to lead to outmigration of skilled personnel from the public sector, fragmentation of risk pools and a growth of power blocs resistant to regulation.5 25 In Uganda, for example, a decline in public financing and public health infrastructure triggered the emigration of many health workers, with a rising number of private sector services, including many illegal clinics and unregistered medical practices.26–28 Private voluntary insurance in ESA largely extends cover for higher income communities and formal sector workforces, with limited pro-poor benefit,29 while a focus on curative personal care in the for-profit sector over population health interventions is reported to exclude poor people and, particularly women, from positive health outcomes.5 While voluntary insurance includes some reimbursement of public and not-for-profit services in a number of ESA countries, only one country, South Africa, is currently working towards pooling tax and non-contributory and contributory financing in a national health insurance covering all members of the population.29

OOP spending below 15% of total health spending has been reported by WHO to reduce the possibilities of catastrophic expenditure, and the consequences for impoverishment and inequity.30 Table 1 shows that in 2018, 11 ESA countries exceeded this level of OOP spending on health. In part, this reflects rising private sector costs. In Kenya, for example, private sector prices were reported to have risen by 20% annually due to the collapse of agreements on pricing guidelines, with the introduction of new co-payments or a reduction in benefits used to manage cost escalation.5 While most ESA countries have policies of free services for particular levels or areas of public sector care, such as maternal and child healthcare, fee charges persist. When public funding does not cover the cost of adequate quality care in public services, informal charges may be levied in the public sector or households may shift to use more costly private services, contributing to rising OOP spending or a fall-out from health services, with negative health consequences, especially among lower income groups.31–33

While the specific implications of a rising share of private spending for health sector preparedness for COVID-19 would need further exploration, a regional review of responses to COVID-19 found the public sector role to be ‘critical and responsible in ensuring a co-ordinated, equitable and comprehensive response across all sectors, and for regulating and accrediting private sector activities.’(Chanda-Kapata, p2)6 The health financing evidence suggests therefore that efforts to mobilise additional private finances for health cannot substitute attention to ensuring adequate public sector health financing, particularly given the size of the financing gap noted earlier for even basic services.

Taxes as a source of health system revenue

Governments collect revenue from taxes on income and profits, social security contributions and from other sources, including taxes on goods and services, payrolls and ownership and transfer of property. Tax revenues can play numerous functions, including redistributing income and wealth, repricing goods and services, and incentivising/disincentivising certain practices, including those that affect health. Tax systems also enable representation of various social groups on the basis of their rights as taxpayers.34 Direct taxes like income tax are levied directly on a taxpayer and adjusted to their characteristics. Indirect taxes are levied on transactions irrespective of the circumstances of the buyer or seller, such as on goods or services.35

The share of tax revenue from different sources for ESA countries is shown in figure 1. In 2018, taxes on goods and services, and taxes on income, profits and capital gains contributed similar shares (46%–47%). Value added tax (VAT) accounted for 27% of total tax revenue, higher than the 22% from taxes on income and the 15% from taxes on profits.36 While figure 1 indicates the variability between ESA countries, the generally lower share of taxes on corporate profits in the region suggests weak tax collection from this source, related in part to global tax rule and profit outflows, discussed later.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Different taxes as a percentage of total tax, ESA countries*, 2019. Source: authors from OECD.36 *Zimbabwe not included due to insufficient data. DRC, Democratic Republic of Congo; ESA, East and Southern African; VAT, value added tax.

Beyond the total share of revenue, different sources of tax revenue need to address key health sector goals, including equity. Table 3 shows the progressiveness of the various sources of health financing for two ESA countries.23 It indicates that direct taxes are the most progressive source of revenue and can generate large funding pools, although they may be affected by economic downturns and be difficult to collect where informal employment is high. Health financing from indirect excise taxes, such as on tobacco and alcohol, is progressive if these goods are consumed to a greater degree by wealthier social groups, although in South Africa, excise taxes and the fuel levy were found to be regressive.37 As shown in table 3, private voluntary health insurance is a regressive form of taxation. Earmarking a surcharge on company or income tax, as applied in Zimbabwe’s ‘AIDS levy’, or a portion of VAT, as applied in Ghana to fund its national insurance, can add significant tax revenue for health.29 38 VAT contributes a significant amount to tax revenue in ESA countries, and may be a progressive source of funds for health where there is a large informal sector, provided it covers commodities consumed more by high-income groups, and if VAT collection thresholds protect small businesses.29

The tax to GDP ratio and tax gap

Despite its importance, the share of tax revenues in GDP in African countries of 16.6%, or the tax: GDP ratio, is one of the lowest in the world. It compares to averages of 21% in Asia Pacific, 22.9% in Latin America and the Caribbean and 33.8% in OECD countries.13 ESA countries have an average tax to GDP ratio of 18.1%.13 Tax to GDP ratios have increased by 1.8 percentage points over the past decade in Africa, primarily driven by increased VAT receipts and a growth in personal income taxes.39 Table 4 shows the wide variation between ESA countries in their tax to GDP ratio from below 10% in Angola and the DRC (countries with high levels of oil and mineral wealth), to ratios above 25% in South Africa, Seychelles, South Africa, Mozambique and Namibia.

Each country has an optimum tax to GDP ratio, referred to as its tax capacity. The sub-Saharan African tax capacity is conservatively estimated to be 20%.40 The gap between this and actual tax capacities represents lost tax revenue. For the ESA region, the average tax capacity of 18%, yields an estimated tax gap of 2% of ESA GDP. With a GDP of US$773 billion in 2018, this indicates a lost tax revenue of US$15 billion, or a tax loss of US$34.2/capita for the 483 million people in the 17 ESA countries.24

Comparing the tax to GDP ratios in table 4 with the share of government spending on health in table 1 suggests that those countries with higher tax capacities also have higher shares of government spending on health. In Lesotho, Mauritius, Namibia, Seychelles and South Africa, tax capacities were greater than 20% and shares of government spending on health also exceeded 10%. There were outliers, although few, such as Mozambique, where tax was 27% GDP but the share of government spending on health only 5.6%, or conversely Botswana, with a lower tax capacity of 13% but a high share of government spending on health at 14.3% in that year.

Losses to ESA tax revenue from global tax rules and IFFs

Several factors have contributed to the observed low tax to GDP ratios. The High Level Panel on IFFs from Africa reported poor governance, weak regulatory structures, harmful tax incentives and double taxation agreements as contributing to lower levels of tax collection.41

Tax losses can also be attributed to a global rule system that assigns tax collection to countries outside the region.

In a separate entity principle, multinational companies that operate under a common group and ownership can treat their branches in different countries as independent entities that work at arms-length. This enables companies to employ strategies to reduce their taxable income, termed ‘base erosion’ and to shift profits to other lower tax countries, both of which affect tax collection. In contrast, a unitary taxation approach allocates the overall profits of multinationals to the different states they operate in. Historically, the formula for country allocations has used the share of payroll costs, sales and physical assets, with continued benefit for high-income countries.12 In 2016, and in response to grievances raised around the inequity in the current global tax rules, the OECD proposed adapting the formula for tax liabilities for some multinationals to be applied on the basis of sales where final consumers are located. This, however, applies only to 25% of the profits of companies with revenues of more than US$21 billion and a profit margin above 10%, affecting less than 100 companies globally.42 Given the high volume of export of African commodities, it biases towards consumption in high-income countries.

Given this limited scope, African finance ministers have raised concerns that developed countries are not listening to the concerns of low/middle-income countries and are not significantly redressing an unfair balance in taxing rights.43 The demand for change towards a fairer global tax system became even more pressing given the significant costs generated by the COVID-19 pandemic and the increased deficits generated by the resultant trade and economic disruption.

Alternative approaches have been proposed. One proposal is for all countries to apply a minimum effective tax rate (METR) to avoid the incentive of shifting declared incomes to low tax countries or tax havens. A proposed METR of 25% is closer to Africa’s average corporation tax rate of 28%, given that an OECD proposal of 15% would still incentivise profit-shifting to OECD countries, given the higher rate in Africa.44

Table 4 shows how much ESA countries could gain by applying an METR of 25%. All ESA countries achieve some level of tax gain, and some, such as South Africa, Angola and Zambia, Lesotho and Mauritius, benefit more than others.45 Using the data in table 4, excluding the three ESA countries for which data were not available, applying a 25% METR in global tax rules would lead the region to gain US$7.2 billion in additional tax collection annually, or US$26.2/capita.

IFFs also contribute to losses in ESA tax revenue, as ‘financial flows that are illicit in origin, transfer or use; that reflect an exchange of value instead of a purely financial transactions; and that cross country borders’ (46:online) African countries are estimated to have lost more than US$50 billion annually over the past 50 years, exceeding the annual inflow of overseas development aid.41 An annual outflow of financial resources from the continent of US$88.6 billion reported in 2020 suggests that IFFs have increased.47 Most (65%) of these IFFs arise from commercial activities, while illegal activities and corruption account for 30% and 5% of IFFs, respectively.41 Commercial IFFs stem from transfer pricing, trade mispricing, misinvoicing of services and intangibles and use of unequal contracts, enabling tax evasion (illegal), aggressive tax avoidance (legal) and illegal export of foreign exchange.41 Legal tax avoidance is also enabled by the global rules system noted earlier, in providing legal loopholes to minimise tax liabilities in African countries.41

Depending on the estimate, base erosions and profit shifting to tax havens alone collectively cost low-income countries between US$167 and US$200 billion annually in lost corporate tax revenue.48 Table 4 shows the estimated losses for ESA countries, drawing on 2017 data.39 The evidence suggests that the ESA region lost US$7.6 billion annually in tax revenue due to global practices of base erosion and profit shifting, or US$124.7/capita, equivalent to 1.6% of the region’s GDP. Given that Seychelles and Mauritius with their low populations contribute to high per capita averages, excluding these two countries suggests an average annual per capita loss from these two commercial tax practices alone of US$13.8. This is a conservative estimate as it excludes losses due to other IFF sources and losses due to limited taxation of natural resource depletion of extractive activities.