Effect of economic uncertainty on public health expenditure in Economic Community of West African States: Implications for sustainable healthcare financing

Abstract This study investigates the dynamic effect of economic uncertainty on public health expenditure in the Economic Community of West African States region. The investigation is motivated by the recent volatilities in the global economy in the face of increasing demand for adequate funding of health systems in the region, and the need to fill the existing gap in the literature. The study employs the panel autoregressive distributed lag model to express the theoretical relationship between public health expenditure per capita, economic uncertainty and population growth rate, and estimates the model parameters using the mean group and the pooled mean group estimators, after accounting for stationarity and cointegration. Results reveal that on the aggregate, economic uncertainty and population growth are significant determinants of per capita health spending in the long run. When the countries are disaggregated by income groups, evidence suggests that in low‐income countries, economic uncertainty is negatively associated with health spending in the short run, while a growing population reduces health spending per capita in the long run. In lower‐middle‐income countries, economic uncertainty increases health spending in the short run, but reduces it in the long run as uncertainty persists, while population growth negatively impacts health spending in the long run. We conclude that the dependence on public funding of the health system in the region appears unsustainable. Thus, health financing policies need to explore alternative funding mechanisms that entrench cost‐sharing between the public and private financiers.


| INTRODUCTION
In developing countries characterized by low levels of investment, income, employment opportunities, life expectancy, and human development index on the one hand, and rising poverty levels on the other hand, public expenditure has become a major policy tool for facilitating the development of a country's human capital (World Health Organization [WHO] 1 ), and for spurring the economy unto the path of growth and sustainable development. The thinking among development practitioners is that increased public spending on healthcare should be a priority for governments at all levels. This thinking is motivated by the benefits of adequate funding of the health system for the poor population on the one hand, and the negative macro and microeconomic consequences of inadequate funding of the system on the other hand.
An optimally funded health system ensures improved health status and human capital; reduced burden of out-of-pocket (OOP) health expenditure for poor households; and facilitates the development of the human capital required for sustainable development. 2 On the other hand, suboptimal public health spending can cause severe socioeconomic consequences for poor households due to high cost of accessing health services from private providers. High cost of healthcare could result in catastrophic OOP health expenditure as households are forced to pay a large share of their income on health services, pushing some into poverty, and others into even deeper poverty than they are already in. Households seeking health services may be forced to borrow money, sometimes at very high-interest rates, or to sell their assets, to pay for health services. The alternative for such households is to forgo health services and live with their illness and suffer the short-and long-term consequences. 3 A major challenge of public health financing in developing countries is the impact of global economic uncertainty on domestic health budget. Economic uncertainty is one of the inevitable features of a globalizing world. The macro and microeconomic impacts of economic uncertainty have been a source of concern to policymakers especially in developing countries that mostly suffer the severe impacts of the phenomenon. There are concerns that reduction in public health spending due to economic uncertainty could affect the poor and vulnerable populations and, in turn, erase the development progress that has been made thus far. Addressing existing poor health outcomes within the context of economic crisis has become a policy priority given the importance of a healthy population in a developing country. 4 Also, understanding how a growing population affects health expenditure is critical for evolving sustainable health financing policies. This is because evidence suggests that a growing population can increase the fiscal burden of health expenditure on the government. [5][6][7] Various studies have attempted to understand how economic uncertainty impact public spending on healthcare. An examination of the impact of economic crisis on healthcare resources in the Eastern Mediterranean countries of WHO reveals that being unemployed and having to spend from OOP are negatively correlated with healthcare expenditure per capita: a 1% rise in unemployment is found to decrease health expenditure per capita by $138, and an increase of poor and vulnerable populations will have to rely on government health facilities to access affordable healthcare, the funding of which may be vulnerable to economic uncertainty, and third, achieving the third goal of the sustainable development goals (SDGs)-good health and wellbeing-is dependent on improved funding of the health system, indicating the need to understand how public health financing is affected by economic uncertainty. Equally, achieving this objective will provide empirical answers to the question: what is the effect of economic uncertainty on public health spending in the ECOWAS region? These provide compelling justifications for policymakers to understand how public spending on healthcare is impacted by economic uncertainty. The study contributes to existing literature on the sustainability of public health financing amid uncertain regional economic outlook by highlighting how public health financing could be impacted by economic uncertainty, as well as the magnitude of such impact.
ECOWAS is a regional economic group comprising of all the 15 countries that make up the West African region, with a mandate of promoting economic integration. The Commission was set up to foster the ideal of collective selfsufficiency for its member states and works to harmonize macroeconomic policies toward achieving regional economic integration. 17 In 2001, African Union heads of state, including those of ECOWAS countries, pledged to allocate at least 15% of their annual government budget to the health sector under the Abuja Declaration. This commitment marks an important initiative in the history of public health financing in the region. 18 However, evidence reveals that between 2010 and 2018, no West African country attained the 15% threshold for public budgetary allocation to the health sector. The top three countries in this regard, Ghana (8.43%), Carbo Verde (8.29%), and Burkina Faso (7.60%), still allocates less than 10% of their total annual budget. Other countries, for example, Guinea (3.05%), Liberia (3.46%), and Guinea-Bissau (3.56%) spend less than 5% of their annual budget. 19 Given this scenario and coupled with the COVID-19-induced global economic uncertainty that has affected the economies of the countries of the region, this study becomes timely and policy-relevant.
As far as we know, no study on this issue has been undertaken in the region. The closest approximation to this study is the work by Aregbeyen and Akpan. 10 However, their study differs from this study in scope and methodology. Their study adopts a microoutlook and considers the effect of economic shock on health expenditure of rural households using the Heckman's selectivity model. Our study focuses on the macro dynamics of economic uncertainty on public health spending. This is important because of the growing demand on the government for increased health budget for a poor, growing population. Methodologically, we adopt the panel autoregressive distributed lag (PARDL) model. This model allows us to determine the dynamic effect of economic uncertainty on public health expenditure in the short and long run, while accounting for group-specific effect.
The rest of the study is structured as follows: immediately following this introduction is Section 2, containing the details of the data and method. In Section 3, we present the estimated results and their interpretations. Section 4 contains the conclusion and policy implications of the study.

| Data sources
This is a macroeconomic panel data study that uses data from the 15 countries in the ECOWAS region for a period of 19 years

| The model
Various theories provide broad framework for understanding the linkage between growth in government expenditure caused by increasing demand for social services. The seminal work of Adolph Wagner-Wagner's hypothesis of increasing public debt-explains that government expenditure is bound to grow in an industrializing economy as the government strives to provide basic infrastructure required to support industrialization. One of the major arguments of the law is that because the goods and services (e.g., healthcare) supplied by the public sector have high-income elasticity of demand, as this elasticity increases, public expenditure must increase proportionally to the increase in income. 20 Musgrave hypothesis attributes the growth in government expenditure to structural adjustments that coincide with the development process of a country. In the initial phase of development, private capital formation is low, while population continues to grow. This leads to reliance on publicly provided goods and services, leading to the burgeoning of public budget. 21 The Brown and Jackson 22 microeconomic theory of public expenditure considers the demand side factors (taste and income) and the supply side factors (tax rate, and cost of production) to determine factors influencing public expenditure growth. The model recognizes the service environment, population growth-leading to increased demand for healthcare services-and changes in the quality of public goods demanded by the median voter as determinants of growth in public expenditure. This study adopts the Brown and Jackson microeconomic theory of public expenditure as the framework for its model, and hypothesis that public health expenditure in the ECOWAS is affected by economic uncertainty.
Various studies have employed the Brown and Jackson framework.
They find population and economic growth as determinants of public expenditure. 10,23,24 The innovation of this study is the introduction of economic uncertainty into the model as one of the determinants of IHEOMA | 3 of 10 public health spending. The PARDL model for consideration is given as: In Equation (1), Χ it , ϒ it , Ζ it are health expenditure per capita, economic uncertainty, and population growth rate, respectively for individual countries. γ i , λ i , ϕ i are the corresponding long-run parameters of the variables, ε it is the disturbance term, i and t represent country and time respectively, while k indicates the optimal lag length. The model is justified as the period of coverage (T = 19) is larger than the number of cross sections (N = 15). 25 The Testing for cointegration is a necessary step to establishing if variables empirically exhibit meaningful long-run relationships. 33 If we do not reject the null hypothesis, we conclude that the variables are devoid of any long-run relationships, in which case we restrict our analysis and interpretation to the short-run estimation. Conversely, if we do not accept the null hypothesis, we reparameterize Equation (1) into an error correction model. Doing so produces a model that incorporates the short and the long-run information regarding the interaction of the variables, as well as the error correction mechanism.
Where all other variables and parameters remain as previously defined. ρ i captures the speed of adjustment to long-run equilibrium.
It is derived as the error term from Equation (1) Equation (3) is the mean equation. The dependent variable (ϒ it ) is economic growth measured by the growth rate of the gross domestic product. C is the constant, while σ it−1 is the error term. After estimating Equation (3), we extract the residuals and use them as the dependent variable in Equation (4). Equation (4) assumes that the squared residual from the preceding equation is a function of its lagged value and a random innovation (ε t−1 2 ). After estimating Equation (4), we extract the series of the random innovation and use that as our ϒ it in Equations (1) and (2).
We use the mean group (MG) and the pooled mean group (PMG) estimators to determine the relationship between the variables. The PMG restricts the long-run estimates to be equal across countries, while allowing them to differ in the short run. The short-run relationship captures country-specific heterogeneity which may arise from the unequal magnitude of economic shock for each country.
Conversely, The MG estimator allows for heterogeneity in the short and long run relationships. 28 The Hausman test is used to select the optimal estimator. By

| Empirical results
In were recorded in LICs and LMICs respectively, and the rates are relatively stable in both groups. Results from the LICs (Table 5)  Note: ***, **, * denote the rejection of the null of a unit root for 99%, 95%, and 90% significant levels respectively.
Abbreviation: IPS, Im-Pesaran-Shin; LLC, Levin-Lin-Chu; LMICs, lower-middle-income countries.  This increased demand is expected to lead to increased government funding of the health system as the government tries to cater for its and in Southeast Asia. 5 The reason for this difference in results is rather intuitive: a growing population leads to increased demand for public health services, especially in poor economies where household income is inadequate to satisfy the household healthcare needs.

T A B L E 3 Panel cointegration tests
Where health expenditure does not grow proportionally to population growth, the size of the public health budget available per individual declines.
The study is limited to the effect of economic uncertainty on public health expenditure due to the availability of comprehensive data set on that aspect of healthcare financing. Further studies may consider other forms of healthcare financing such as private health expenditure and external health expenditure. Also, investigation of the effect of economic uncertainty on household health spending may be an innovative research area. Other factors affecting health expenditure other than economic uncertainty may also be explored in subsequent research.

TRANSPARENCY STATEMENT
The lead author (manuscript guarantor) affirms that this manuscript is an honest, accurate, and transparent account of the study being reported; that no important aspects of the study have been omitted; and that any discrepancies from the study as planned (and, if relevant, registered) have been explained.