Revenue-raising potential for universal health coverage in Benin, Mali, Mozambique and Togo

Abstract Increasing overall fiscal space is important for the health sector due to the centrality of public financing to make progress towards universal health coverage. One strategy is to mobilize additional government revenues through new taxes or increased tax rates on goods and services. We illustrate how countries can assess the feasibility and quantitative potential of different revenue-raising mechanisms. We review and synthesize the processes and results from country assessments in Benin, Mali, Mozambique and Togo. The studies analysed new taxes or increased taxes on airplane tickets, phone calls, alcoholic drinks, tourism services, financial transactions, lottery tickets, vehicles and the extractive industries. Study teams in each country assessed the feasibility of new revenue-raising mechanisms using six qualitative criteria. The quantitative potential of these mechanisms was estimated by defining different scenarios and setting assumptions. Consultations with stakeholders at the start of the process served to select the revenue-raising mechanisms to study and later to discuss findings and options. Exploring feasibility was essential, as this helped rule out options that appeared promising from the quantitative assessment. Stakeholders rated stability and sustainability positive for most mechanisms, but political feasibility was a key issue throughout. The estimated additional revenues through new revenue-raising mechanisms ranged from 0.47–1.62% as a share of general government expenditure in the four countries. Overall, the revenue raised through these mechanisms was small. Countries are advised to consider multiple strategies to expand fiscal space for health.


Introduction
Countries may need to raise additional funds to progress towards universal health coverage (UHC). This implies increasing the fiscal space for health. Fiscal space has been defined as "the ability of governments to increase spending for the sector without jeopardizing the government's long-term solvency or crowding out expenditure in other sectors needed to achieve other development objectives. " 1 Fiscal space for health can be expanded in several ways: general economic growth in a country; increased state or tax revenues and improved tax collection; an increased proportion of government spending on health; and improved efficiency in the use of funds. 1,2 Mobilizing additional tax revenues can be done by introducing new taxes or increasing existing tax levels. Imposing taxes on specific products and services to increase general government revenue has also gained attention through the World Health Report 2010. 3 Countries' interest in resource expansion for health is increasingly important in the light of decreasing levels of funding by global health initiatives to low-and middleincome countries. 4 Importantly, raising additional revenue for health needs to be examined within the context of overall government revenues, of which health is only one component. The objective to increase fiscal space for health does not necessarily require new revenues to be earmarked for the health sector, although some countries do so. Instead, the aim is to increase overall government revenues and augment the share going to health. 2 While a mix of strategies may be needed to expand fiscal space, we focus in this paper on mechanisms for raising additional government revenue. We illustrate how countries can assess the feasibility and quantitative potential of the mechanisms. To do this, we review and synthesize such processes and results from four country studies in Benin, Mali, Mozambique and Togo. [5][6][7][8] The studies were part of the countries' efforts to develop strategies to expand UHC. Table 1 summarizes key demographic, health and health coverage indicators of the four countries. The data show that there is still a long way to go towards UHC. For example, the UHC service index which measures coverage of essential health services ranged from 32 to 42 across the four countries, compared with above 70 in Organisation for Economic Cooperation and Development countries. 11 The share of the population working in the informal sector is high (Table 1). Currently, people rely largely on underfunded, government health services. Benin has begun to build up a national insurance scheme in which funds from the government budget would be used to finance the health coverage of the very poorest people and to partially subsidize poor people, while higher economic groups would make contributions. 13 In Mali, the parliament approved a law in 2018 on a national universal health insurance scheme, but implementation has not yet started. The idea is to use state budget transfers to subsidize the contributions of vulnerable Raising revenue for health Inke Mathauer et al.

Context of country studies
and poor population groups in the informal economy. The health ministry has projected the funds needed to provide these subsidies, with a core assumption being an increased budget for the health sector. 14 However, the precise source of revenue and which additional revenueraising mechanisms will be applied has not yet been decided. In Togo, the health ministry is in the process of finalizing a national health financing strategy. The existing mandatory health insurance scheme is still limited to current and retired civil servants and their family members, and covers 4% of the population in 2019. 15 Contributions are paid by the civil servants and their employer (government agencies). Hence, a core question is how to expand coverage to the whole population. Technical debates currently focus around the idea of using budget transfers to cover people in the informal economy. 16 Benin, Mali and Togo are members of the West African Economic and Monetary Community. The Community provides a harmonized tax framework, which sets a limit on specific taxes (tobacco products and alcoholic drinks, for instance) and has harmonized taxation rules for certain sectors, such as banking and aviation. 17 In Mozambique, the government has developed a health financing strategy, which is currently subject to approval from ministries. In this strategy, the aim is to define various mechanisms to raise financial resources to enhance fiscal space. 18 Mozambique is part of the Southern African Development Community, which also seeks to harmonize certain tax rates among member countries. 19 Table 2 presents some key health expenditure indicators and reveals that domestic general government health expenditure as a share of current health expenditure is low in the three west African countries (ranging from 20.0% to 31.1 %). In Mozambique, the figure is higher (53.3%), but its per capita current health expenditure is also much lower than in the other three countries. The priority given to health and hence the budget allocation to health (which includes domestic general government health expenditure and the external funds flowing into the health budget) as a share of general government expenditure is still rather low. 20 Likewise, general government expenditure as a share of gross domestic product (GDP) is still low for Benin and Mali (21.3%   21 Global cross-country evidence shows that the absolute level of public spending matters and a systematic improvement in UHC performance, in particular a lower incidence of catastrophic health expenditure, is observed when public spending on health increases. 22,23 Thus, the four countries' UHC expansion efforts would benefit from more revenues through an overall increased government budget and a higher share of this going to health.

Illustrating the assessment approach
We outline a four-step method and process that was applied to assess new revenue-raising mechanisms in the four country studies. Each country study was part of the technical and policy advisory support process that was requested from the World Health Organization (WHO). Each country study team consisted of a national and international consultant, from among the authors with this specific expertise, accompanied by the country's health ministry and WHO country office and headquarters staff.

Multistakeholder consultation
The first step was a multistakeholder consultation in each country that served to pre-select the new revenue-raising mechanisms to be explored in detail. A wide range of stakeholders participated in a one-day meeting: representatives from ministries of health, finance, tourism services and infrastructure; civil society organizations; development partners; and the private sector. Following the same format and approach in each country, study teams presented a range of revenue-raising mechanisms, with their advantages and disadvantages, based on evidence from the literature. Small group and final plenary discussions of what stakeholders considered useful resulted in a shortlist. The list was screened for a final selection of four to five revenue-raising mechanisms to be explored in depth (Box 1).

Feasibility analysis
In the second step, each country team conducted a detailed qualitative analysis of the feasibility of the selected mechanisms. This started with a literature and document review, which informed the subsequent data collection process. A series of semi-structured interviews were held with key stakeholders from government agencies, the private sector and development partners. The interviews provided insights into current taxation mechanisms and rates in the respective sectors, the feasibility of the mechanisms explored, and potential challenges, such as whether stakeholders would support or resist the introduction of a new revenue-raising mechanism. This qualitative analysis was guided by six criteria looking at various aspects of feasibility (Box 2). The criteria were developed during the first country study in Togo 5 and applied in the other three studies. We graded the criteria from very weak to very strong based on the data from stakeholders' discussions and interviews.

Quantitative analysis
The third step was the quantitative analysis. The country teams collected data from country statistics and global databases, such as World Bank development indicators, the International Monetary Fund's world economic outlook indicators and WHO global health expenditure data. This step also served to set assumptions and projection variables to estimate potential revenues for different scenarios, for a defined projection period which was determined at the stakeholder meetings. Box 3 illustrates the approach to estimating revenues, taking the example of a tax on airplane tickets in Togo. Different high and low scenarios were specified for each mechanism to estimate potential revenues for the defined period (Table 3). For example, a high scenario was based on a higher tax rate or assumptions of higher increases in the consumption of a product or a higher growth rate over the projection period.

Box 1. Revenue-raising options discussed at stakeholder consultations and selected for the in-depth analysis in the country studied Benin
Discussion of taxes on: airplane tickets, financial transactions, alcoholic drinks, tobacco products, public contracts, imported vehicles, petroleum products, extractive industries, large companies, real estate property, luxury products, companies with large volume of pollution, household garbage, mobile phones, large cars, lotteries and gambling, health insurance contracts, pharmaceutical companies of branded medicines, voluntary diaspora contributions, or an increase of VAT and of traffic violation fees. Selected taxes for in-depth analysis on: alcoholic drinks, airplane tickets, telephone (mobile), financial transactions and national lottery.

Mali
Discussion of taxes on: airplane tickets, visa applications, alcoholic drinks, tobacco products, public contracts, hydrocarbon, hotel nights, extractive industries, sugar-sweetened drinks, real estate property, transport companies, companies with large volume of pollution, earnings of ministers and deputies, mobile phone calls, livestock exports, lotteries and gambling, pharmaceutical companies of branded medicines, voluntary diaspora contributions, road tolls, financial transactions; or an increase of municipal taxes and of VAT. Selected taxes for in-depth analysis on: alcoholic drinks, airplane tickets, telephone (mobile and fixed), financial transactions and extractive industries.

Mozambique
Discussion of taxes on: alcoholic drinks, tourism services, vehicles, extractive industries, private clinics, forestry and wildlife activities. Selected taxes for in-depth analysis on: alcoholic drinks, tourism services, vehicles and extractive industries.

Togo
Discussion of taxes on: airplane tickets, financial transactions, alcoholic drinks, tobacco products, public contracts, imported vehicles, petroleum products, extractive industries, large companies, real estate property, luxury products, companies with large volume of pollution, household garbage, mobile phones, large cars, lotteries and gambling, health insurance contracts, pharmaceutical companies of branded medicines, voluntary diaspora contributions, or an increase of VAT and of traffic violation fees. Selected taxes for in-depth analysis on: alcoholic drinks, airplane tickets, telephone (mobile and fixed), financial transactions and extractive industries.

Stakeholder feedback discussions
In the fourth and last step of this process the country teams reported back the results of the qualitative and quantitative analysis to all stakeholders and decision-makers at a workshop to receive feedback on the suggestions. The workshop also served to build ownership on the conclusions and translate the analysis into an agreed way forward for policy discussions and decisions on next practical steps, also in relation to the development or the implementation of the health financing strategy.

Illustrations of country findings
The list of mechanisms selected for the in-depth studies and the feasibility issues expressed by stakeholders were similar in the three West African countries (Table 4). Stability and sustainability were rated positive for most mechanisms, except for a tax on the extractive industries and national lottery tickets. Stakeholders thought that a new tax on remittances might raise equity concerns due to potentially negative impacts on lower income groups. Tax differentiations between consumer goods (wines and spirits versus beer in the case of a tax on alcoholic drinks) and consumer groups (business versus economy passengers in the case of a tax on airplane tickets) can make the tax more progressive. Political feasibility seemed to be an issue for nearly all the mechanisms assessed. Taking all feasibility criteria into consideration, new taxes or increased tax levels on alcoholic drinks, airplane tickets and telephone calls received the most positive ratings in the feasibility assessment. Taxes on national lottery tickets, financial transactions and the extractive industries were rated as less acceptable. Stakeholders argued that the financial sector and extractive industries are emerging and need to attract investors and the political situation around the extractive industries was still unclear.
For Mozambique, stakeholders assessed most of the studied mechanisms positively regarding sustainability, progressivity and potential trade-offs, but rated political feasibility lower, due to the likely competing interests of different ministries (Table 5). Moreover, administrative efficiency was a concern for taxes on the extractive industries, since the set-up and running costs of the tax are expected to be high and technical capacity to be weak. Overall, stakeholders rated new taxes on alcoholic drinks and on tourism services as more promising. Table 6 illustrates the quantitative potential for raising revenue of the low-scenario and high-scenario cases (i.e. the combination of all low-scenario settings for each mechanism, or of all high-scenario settings respectively), as well as of the basket of revenue-raising mechanisms that were proposed for further policy consideration ( Table 3). The range of estimated additional revenues, as a share of general government expenditure, that could be mobilized from this suggested basket of revenueraising mechanisms were 0.47-1.62% across the four countries, or 0.52-2.88% for the high-scenario case. [5][6][7][8]

Policy lessons and key issues
The results from both the qualitative and quantitative assessments showed that the proposed new revenue-raising mechanisms could be feasible options for increasing domestic revenues. The estimated additional revenues as a share of general government expenditure from the suggested basket of revenue-raising Box 2. Feasibility criteria and related key questions for the qualitative assessment of revenue-raising mechanisms

Political feasibility
Is there political will for this funding mechanism, or does it create reluctance at the political level (whether from government or civil society)?

Sustainability
Would the mechanism be applicable in the long term?

Stability
Would revenues be stable over time?

Progressivity (equity in financing)
Would financially better-off people likely contribute with a larger proportion of their income than poorer people?

Administrative efficiency
Are institutional and operational arrangements in place to implement the financing mechanism? What would be the risks of fraud and corruption and how could these be reduced?

Other possible effects
Which (positive or negative) effects would this revenue-raising mechanism have on the supply and demand of particular goods and services?

Box 3. Example of scenario definitions and assumptions set to estimate revenues from an airplane ticket tax in Togo
Projection period: 10 years Definition of different taxation scenarios: • scenario 1: taxing only passengers going abroad; distinction of taxes between economy class and business class; • scenario 2: scenario above plus taxing arrival passengers; • scenario 3: scenario 2 plus taxing transit passengers.
Setting of assumptions over the projection period for: economic growth, demand elasticity and inflation rates; share of business-class or first-class versus economy-class passengers. Projection of the number of passengers departing from, in transit and arriving in the country, in business and economy class, over the projection period, based on the above assumptions. Calculation of potential revenues, using the above scenarios and assumptions, was done using the following formula: revenues (in national currency) = tax rate (%) x tax base (in national currency) with the tax base calculated as: number of services or number of consumed products multiplied by the elasticity factor, projected over the number of years with estimated growth rate and inflation rate for each year.
Note: Explanations on more detailed formulas can be found in country studies [5][6][7][8]   No scenario defined due to lack of data c Due to lack of accurate data and simplicity, it was assumed that circumstances would remain the same as under the low scenario. Notes: A new tax refers to introducing a new type of tax, independent of whether another type of tax (for example a value added tax) existed on the same product or service. An increased tax rate refers to an existing tax that is raised. Source: Based on country studies. [5][6][7][8]   options are rather small. Nevertheless, even a small increase in revenue is valuable. This finding is in line with the evidence from a recent WHO review that reiterated the importance of increasing fiscal space through new general revenue-raising mechanisms in combination with other strategies to expand the fiscal space for health. 2 Consideration of various limitations and implementation issues is important. Unavailable or inaccurate data made it impossible to adequately estimate potential revenues for a few mechanisms, particularly for a tax on the extractive industries. There also remains uncertainty about how realistic the assumptions are. These factors affect the strength of the projections. Moreover, it is unlikely that countries would imple-ment the full basket of mechanisms under consideration. Also, the estimates do not consider existing shortcomings in tax administration and collection (including tax evasion, smuggling and the informal economy), which would reduce the estimates of revenues raised.
The stakeholder consultations and interviews revealed that some sectors seemed more attractive than others for the introduction of new revenue-raising mechanisms. This was the case for a new or an increased tax on airplane tickets, telephone calls and (imported) alcoholic beverages in Benin, Mali and Togo. In Mozambique, new taxes on tourism services, alcoholic drinks and the extractive industries and an increased tax on vehicles were considered as possible options. This attractiveness may also relate to the fact that some of these taxes are already in place in other countries in the region and worldwide, and will be paid by a large share of people. For example, Gabon is well known for collecting a tax on the turnover of mobile phone companies. 26 More than half of the funding for the international drug purchasing facility Unitaid comes from a tax on airline tickets levied by 10 countries . 27 Also, nearly all countries globally already have an excise tax on alcoholic drinks, although few adjust this for inflation. 28 Moreover, most countries worldwide have a tax on tobacco products and although these taxes are mostly rather low, 106 countries have increased their tobacco excise taxes since 2007, after the Framework Convention for Tobacco Control was ratified. 29 Price elasticity of demand for cars is fairly rigid. No effective and efficient alternative means of (public) transport is in place (+) Already annually collected and in place for the lifetime of natural resources (+)

Stability
Growing industry (+) Growing industry and competitive environment (+) No major fluctuations, at least for light and heavy vehicles in the short and medium term (+) Revenues depend on fluctuations of international commodity prices, but industries overall are growing (+ -) Progressivity With a high level of current smuggling, the burden of a new levy would likely affect the formal sector (+ -) The burden of the levy would increase with the price of accommodation (+ +) The levy would be mostly incurred by vehicle owners who can afford to purchase and maintain a vehicle (+) The tax burden of different income groups would not be affected through this earmarking Administrative efficiency Mechanisms to collect taxes are already in place (++)

No information available
Running costs would be high. Building technical capacity will be crucial (--) No mechanisms are in place. Running costs would be high.
Inter-ministerial management committee is required (--) Other possible effects and trade-offs Potential to reduce alcohol consumption, which increases the health status of the population (+) Supply side will likely be challenged to provide better services (+) No anticipated side-effects. Increase in the statutory vehicle tax is unlikely to substantially reduce demand for vehicles (+) Calls for improved and transparent financial management (+ -) Note: (--) very weak; (-) rather weak; (+ -) neutral; (+) strong; (+ +) very strong. Source: Based on country study. 8 Raising revenue for health Inke Mathauer et al.
The countr y studies fur ther demonstrated that exploring the feasibility of new mechanisms is essential, as it may rule out some of the options that appear promising from the quantitative assessments. For example, country stakeholders considered taxing financial transactions and the extractive industries (in Togo and Mali) as not currently feasible. Also, the studies revealed that a feasibility assessment needs to go beyond national borders to consider the role of sub-regional regulations, such as from the West African Economic and Monetary Community for the three West African countries. 17 In terms of the process, the country studies confirmed that a wide range of stakeholders and decision-makers need to be included from the very beginning, to create a mutual understanding of the role of new revenue-raising mechanisms, with an ultimate aim of increasing funds for the health sector for progress towards UHC. While finance ministries will lead such discussions, health ministries can contribute in a constructive way to this dialogue. A set of arguments for ministries of health to use in this dialogue have been suggested by other researchers. 30 The consultation process also allows for raising new considerations for the development of health financing strategies. Moreover, discussions around fiscal space enabled better exchange on health financing with the finance ministry and other ministries and fostered collaborations, as is found by other reseachers. 2 Finally, it is important to carefully assess whether and if so, when, to bring up the issue of earmarking for health into these discussions in order not to affect the health financing and domestic revenue-raising policy dialogue. International evidence points to the fact that earmarking for health may raise additional resources, but this may be offset by reducing discretionary budget allocations, resulting in little if any overall increased fiscal space for health. 31,32 However, from the perspective of finance ministries, tying the messaging and advocacy for a specific tax increase to the health sector may be preferable, as it may increase acceptability by the public.

Conclusion
Discussions on health financing reforms for UHC are ongoing in the four studied countries and so is the process of reflection about new revenue-raising strategies. As in other countries, these are multi-year processes of political negotiations and decisions on new revenue-raising mechanisms remain to be reported. This type of work, however, can trigger or further inform such policy discussions.
In summary, new revenue-raising mechanisms remain a topical subject, as countries seek to estimate the potential of new revenue-raising mechanisms. With a rising burden of noncommunicable diseases, so-called health taxes (on products high in saturated fat, trans-fatty acids, sugar or salt) receive increasing attention, similar to so-called sin taxes (on tobacco products and alcoholic drinks). However, it needs to be emphasized that the primary rationale of such taxes is to reduce the consumption of products with harmful health consequences. Increasing general government revenues is only a secondary objective. 33  25 b In Benin, only a high scenario was calculated. c Estimates of the basket of mechanisms proposed for policy consideration, listed in Table 3. d For Mali and Togo, no data were available to project revenues for a new tax on the extractive industries. Sources: Based on the results of country studies. [5][6][7][8] Total amounts of revenues per high, low and proposed scenario cases were translated into shares as of general government expenditure and GDP.

Policy & practice
Raising revenue for health Inke Mathauer et al.
For future initiatives and studies, there are several key messages. First, whatever the source of additional revenue, in principle such new revenueraising mechanisms should flow into the general government budget rather than being ring-fenced for a specific sector or disease programme. Second, more attention is needed on how to improve tax collection, which is also part of increasing revenues. Importantly, various publications suggest that improved tax collection is one of the most effective strategies to increase government revenues. 2,34,35 Finally, it is important to remember that new revenue-raising mechanisms represent only one of several strategies to expand fiscal space for health and a combination of strategies is needed. While a health financing strategy highlights the need for additional revenues going to health, overall government revenue-raising must be distinguished from the question of health financing for UHC. ■ Raising revenue for health Inke Mathauer et al.