The Impact of Health Expenditure on Maternal and Child Mortality in African Countries

he world has experienced a decline in maternal and child mortality in the last few decades, increase in global health expenditure over the years was the major factor responsible for that. But in Africa and other developing countries, the rate of maternal and child mortality has been soaring. The objective of this paper is to investigate the impact of health expenditure on maternal and child mortality in African countries at different income level. On this note, the paper was based on the panel data covering the period of 2000 to 2017 for 39 African countries. After a cross sectional dependence test and panel unit root test, fixed effect results showed that, government health expenditure has no significant impact on reducing child mortality in African countries at all income level. Whereas, it has an impact on reducing the maternal mortality in low-income and lower middle income countries. It was also found that Private health expenditure has significant impact on reducing child and maternal mortality at all income level. Based on this, the paper recommends that governments at different level especially from low-income and lower middle income countries should provide quality health care services and regular immunization for women and children in order to reduce maternal and child mortality in Africa region.


Introduction
Over the years, economic theories have identified human capital as basis for economic growth and development at macro and micro levels (Novignon & Lawanson, 2017). Health capital contributes to the national productivity and increases the length of good time available for market and non-market activities (Grossman, 1972;Muurinen, 1982). For this reason, improving health capital remains a top priority for a number of countries around the world.
Physical capital was considered the most important input for the production of output before the postulation of human capital theory, which was first proposed in the 1960s. From that time, health and education of the labor force were seen as critical elements in the output. Similarly, human healthy time is essential for both market and non-market activities. In the same vein, health care finance is essential in preserving human capital as well as promoting healthy time to achieve a looming growth and development in the economy. Impact of health-care spending can be measured in terms of health outcomes like maternal and child mortality; however, it was also found that population health is influenced by a variety of factors such as social, political, and economic conditions (Kilanko, 2019).
The literature showed that improvements were recorded in some countries around the world in health outcomes (HO) during the last few decades. This was because of the rising healthcare expenditure (HE). In actual terms, global HE per capita has increased, from US$587 to US$1299 in 2000 and 2015 respectively (WHO, 2012). According to the report by World Health Organization (WHO), since 1990 the global under-five mortality rate (U5MR) has declined by 49%, while the maternal mortality ratio has decreased by 45%, and life expectancy at birth (LFE) has grown from 64 to 71 years (WHO, 2013). However, Poullier et al. (2019) pointed out that health spending has received less attention in government budgets in most African and other developing nations with 10% contribution to global health expenditures. This is despite the effort of Abuja Declaration of 2001, which aims to improve public health spending in African countries by allocating 15% of their annual budget to the health sector. Literature showed that few countries in the region are close to meeting the 15 percent target (Tandon & Cashin, 2016). Only six countries have achieved this goal in 2011 after ten years of the declaration, which include Rwanda (23 percent), Liberia (18.9 percent), Malawi (18.5 percent), Madagascar (15.5 percent), Togo (15.4 percent) and Zambia (16.0 percent) (World Bank, 2016).

| International Review of Social Sciences Research, Volume 2 Issue 3
Although the meaning of health spending varies across countries, Poullier et al. (2012) developed a model that classify the overall health spending, which includes total health spending calculated by adding public and private spending on all health-related goods and services. Social security contributions, various forms of taxation to various parts of government, and external sources, such as grants and loans, are typically used to fund governmental expenditures. Private outlays, on the other hand, include private insurance premiums and prepaid plans, mandated enterprise health expenditure, health expenditure through non-profit health services, and direct payments or out of pocket (OOP) expenditure on health goods, such as co-payments and direct payments by uninsured individuals.
In some African countries, public health spending is mostly funded through grants and loans (WHO, 2010), which may be attributed to ineffective taxation and social security systems.
This explains the region's high levels of private health spending as well as the region's high degree of poverty and increasing catastrophic (out-of-pocket) health spending (World Bank, 2016). This could also explain the region's poor health infrastructure and personnel, as little resources are allocated to health-related infrastructure provision and maintenance (WHO, 2010).
Furthermore, majority of nations in the region have inadequate public health insurance systems, with most health-care systems reliant on ability-to-pay (Ataguba & Goudge, 2012).
A country's health system is said to be effective if its citizens' health status and results are better than those of a similar country with equivalent health-care resources. Infant, underfive, and maternal death rates are major health outcomes on which this study focuses. The infant mortality rate is the number of newborns (under one-year-old) who die per 1000 live births in a particular year. The rate of neonatal and post-neonatal mortality is included in infant mortality.
The neonatal mortality rate refers to deaths in the first four weeks of life, which are usually caused by difficulties during pregnancy, such as congenital abnormalities and low birth weight, as well as issues during delivery, such as birth injuries and asphyxia, and problems after delivery, such as measles and other infections. The rate of post-neonatal death is linked to maternal and some variables such as poverty, insufficient health care, infectious illnesses, and traumas (UNICEF, 2011). Under five mortality rate is the probability that a child will die between birth and exactly five years of age per 1000 live births. Diarrhea, malaria, pneumonia, and other illnesses are among the leading causes of death. The majority of these disorders are caused by poverty, civil war, and malnutrition, all of which are prevalent in the African continent. The likelihood that a woman may die during childbirth is known as the maternal mortality rate. This is calculated as the number of female deaths per 100,000 live births from any cause related to pregnancy and childbirth. Pregnancy-related mortality account for more than half a million deaths each year, with the majority of these deaths occurring in developing countries (WHO, 2012). This is mostly due to lack of access to emergency care, inadequate health care, and family planning procedures.
The infant under-five, and maternal mortality rates are serious concerns, particularly in developing countries, and have drawn the attention of various stakeholders and researchers. The burden of mortality and morbidity among the poor is commonly regarded as one of the world's most pressing public health issues, and it has become a significant worldwide concern. This was expressed in the Millennium Development Goals, and is now mirrored in the 2030 Agenda for Sustainable Development Goals. As a result, UNICEF and the World Bank have made child mortality reduction a priority in their future programs (UNCEF, 2019). According to UNICEF, (2019) over 29,000 children under the age of five die every day, or 21 every minute, primarily due to preventable causes. The majority of these deaths occur in poor nations such as Nigeria, Ghana, and Mali.
The main objective of this paper is to investigate the impact of health expenditure on maternal and child mortality in selected African countries at different income level by disaggregating health care spending into public and private. This paper contributes to the existing literature by examining the link between health expenditure and health outcome in African countries at different income level unlike previous studies such as Anyanwu and Erhijakporet (2015) who concentrated their research only in sub-Saharan African countries, Novignon, Olakojo and Nonvignon (2012) in East Africa, Zounkifirou et al. (2021) in India, and other researchers in specific countries. This study considered 39 African countries based on World Bank classification. The classification of the countries based on income level was given by World Bank upon which the analysis was carried out.

Literature Review
Literature showed that the relationship between health expenditures and health status has variety of outcomes. The health production function model (Wang, 2008) and the human capital model (Wang, 2008) are two theoretical frameworks that are frequently used in making analysis 24 | International Review of Social Sciences Research, Volume 2 Issue 3 in this area. Health spending is one of the most important aspects in the creation of health care.
According Musgrove et al. (1996), on the health expenditure and health outcome nexus, health expenditure has no impact on health outcomes. Similarly, Filmer and Pritchett (2000)  A study by Rana and Goli (2017) covered the period of 1995 to 2013 to evaluate how public and private health spending affect child mortality in 146 developing countries. Random effects model results found that total health care spending has a considerable negative impact on child mortality. The study has linked public spending to lower child mortality while private spending remained unlinked but serves as a supplement to public spending on child care.
According to the study, spending on health care by the government reduces the risk of child death in both absolute and relative terms. Another study used data from 1990 to 2010 to examine the impact of maternal and child health outcomes on economic growth in 180 higher and lower middle income countries. The study found that in 105 nations, under-five mortality and economic growth have a bi-directional association, while maternal mortality and economic growth have a bi-directional relationship in 68 countries (Amiri & Gerdtham, 2013).
Furthermore, using data of 133 low and middle-income countries covering the period of 1995, 2000, 2005, and 2006, Farag et al. (2013 investigated the relationship between health expenditure, health outcomes (infant mortality and child mortality), and the impact of governance. The fixed effect results showed that health spending has a significant effect on reducing infant and under-5 child mortality. The study also discovered that government health spending has a significant effect on reducing infant and child mortality, and the size of the coefficient is dependent on the country's level of good governance, indicating that good governance increases the effectiveness of health services. Evidence from the COVID-19 pandemic, Luong (2020) investigated the influence of health expenditure on health outcomes.
The most recent report of the European Center for Disease Prevention and Control (2022)  Sub-Saharan African nations by Novignon and Lawanson (2017) found a negative and substantial association between health expenditure and child health outcomes, with elasticities of -0.11 for infant mortality, -0.15 for under-five mortality, and -0.08 for under-five mortality (neonatal mortality). The cost of public health care was determined to be significantly higher than the cost of private health care. The fixed and random factors, as well as the time frame 1995-2011, were used to estimate positive and substantial lagged effects between health spending and child health.

Methodology
The theoretical framework for this research is a model developed by Michal Grossman (1972) to examine the impact of health expenditure on maternal and child mortality in African countries. The key premise of the Grossman model is that health may be considered as a longterm investment that yields a healthy time output. Individuals are thought to inherit a starting stock of health that depreciates with age and can be improved through investment. In this approach, the "shadow price" of health is influenced by a variety of factors other than the cost of medical care (Grossman, 1972). It is shown that the shadow price rises with age if the rate of depreciation on the stock of health rises over the life cycle and falls with education if more educated people are more efficient producers of health. Of particular importance is the conclusion that, under certain conditions, an increase in the shadow price may simultaneously reduce the quantity of health demanded and increase the quantity of medical care demanded.
The model structure has been derived from the Grossman demand for health model (Grossman, 1972  Any change in these variables affects the net investment in health stock. Medical care, however, being the most important market good component of the gross investment function, is related with prices and expenses. As a result, when all other factors are held constant, higher medical care usage is associated with higher health care expenditure (HE), and vice versa.
HEit is the public and private health expenditure and Xt is a vector of other factors influencing health status or outcomes at time, i implies individual countries.

3.1.Econometric Techniques a. Cross sectional dependence Test
Cross-sectional dependence is one of the key problems with panel data analysis. As a result, a test of cross-sectional dependence is critical in deciding which econometric technique to use in the research. Cross-section dependence occurs when the error terms of nearby units Where pij denotes a correlation between the errors. H0: no cross-sectional reliance is the null hypothesis, while H1: cross-sectional dependence is the alternative hypothesis.

Panel unit root test
The variables' stationarity was investigated. Because of the consideration of crosssectional panel data, the study employed the cross-sectional augmented Dickey-Fuller (CADF) test as described by Pesaran (2007) under the following equation.
Where, ̅ t-1 and ∆ ̅ −1 indicate the cross-sectional averages of lagged levels and first difference individual series respectively.

Panel data Regression
The impact of health expenditure on maternal and child mortality in African countries was analyzed at each income level using fixed effect model. The country fixed effects model used in this study is the following: Where HOit implies the dependent variables (maternal and child mortality), HE is the private and public health expenditure, Xit is the other explanatory variables used in the study. i represents the country identifier, i=1, ..., 39; t represents the time identifier, t=1, ...,17; is the intercept of the fixed effects model; is the error term of mean equal to 0.

Data and Variables used
The data for this study are annual for 39 countries for the period of 2000 to 2017. The source of the data set is the World Development Indicators (www.ourworldindata.com). It was analyzed using STATA 13 and EVIEWS 11.
The empirical rationale for selecting variables is mainly on the basis of the data availability, past and existing literature. The data on maternal and child mortality were accessed as the health outcomes following Chirwa (2019) The study used maternal and child mortality as dependent variables while immunization, urbanization, incidence of malaria, adolescence fertility, government and private health expenditure were used as independent variables. Immunization rate is used as a proxy to measure the effect of the use of preventive health care services on health outcomes such as child and maternal mortality and increase in immunization is expected to reduce maternal and child mortality. Urban population as a measure of urbanization shows the percentages of population leaving in urban areas, the coefficient is expected to be negative because increase in urbanization lead to increase in health facilities, which will improve health outcomes. Incidence of malaria increases child and maternal mortality, which is expected to be positively related to health 30 | International Review of Social Sciences Research, Volume 2 Issue 3 outcomes. Marriages of women during the adolescent period also affect the physical growth of their children. Domestic general government health expenditure as a percentage of Gross Domestic Product is used and private health expenditure as a percentage of current health expenditure are expected to reduce maternal and child mortality.

Cross sectional Dependence Test
The results of four cross-sectional dependence tests are listed in table 1. All variables, with the exception of malaria incidence (for which values are unavailable), confirmed the significance in all four tests. As a result, the null hypothesis of cross sectional independence is rejected at 1% for all variables, indicating that cross sectional dependence exists between the series. It is crucial to analyze the stationarity of the variables included when cross sectional dependence exists.

Panel unit root Test
Due to the presence of cross-sectional dependence, a cross sectional ADF unit root was used utilizing the methods of Pesaran (2007). The results in Table 3 revealed that maternal mortality, child mortality, immunization, and malaria incidence are all stationary at level when using the constant while government health spending, private health spending, adolescent fertility, and immunization are all stationary at first difference.   that child mortality is not significantly explained by public health expenditure. Thus, increase in private health expenditure lowers child mortality rate in Malawi. This is also a result of the prevalence of corruption that worsen the health outcomes.

Regression result
Furthermore, adolescence fertility, immunization, and incidence of malaria have significant impact on reducing child mortality at all income levels. Increase in urbanization reduces child mortality at all income levels except in lower income countries. This implies that urbanization helps provide available infrastructure and ways of tackling diseases that help society remain healthy, leading to reduction in infant and child mortality to the barest minimum.
The insignificant value of urbanization in low income countries has do with the absence of basic infrastructure and health facilities resulting to increase in child mortality. This result is similar to the finding of Anyanwu and Erhijakpor (2009), Novignon et al. (2012), and Kilanko (2019).  Table 5 shows that government health expenditure has significant impact in reducing maternal mortality in low income and lower middle income countries while in high income countries has no significant impact on maternal mortality. A one percent increase in government health expenditure will decrease material mortality by 25.95 and 21.24 per 100000 live birth in lower and lower middle income countries, respectively. Private health expenditure has 34 | International Review of Social Sciences Research, Volume 2 Issue 3 significant impact on reducing maternal mortality only in upper income countries but is statistically insignificant. This result is in line with the findings of Akinci et al. (2014) and Novignon (2017). Furthermore, adolescent fertility, immunization, and urbanization are statistically significant in reducing maternal mortality in both low incomes, lower middle income, and higher income countries. Emmanuel (2018) obtained the same result that urbanization reduces maternal mortality in Nigeria.
Private health expenditure has more impact on reducing child and maternal mortality in lower middle income and upper income countries in Africa. This is a result of high out-of-pocket private health expenditure that becomes prominent in the region and low government health spending. Given the percentage of public health expenditure spent every year by these African countries, the findings show that this expenditure has no impact in reducing child mortality at all income levels. This may be attributed to quality of institutions and governance in these countries.

Conclusion and Recommendation
A number of research examined the relationship between health expenditure and health outcomes but arrived at different conclusion. In this study, the impact of health expenditure on maternal and child mortality in African countries at different income level using a panel data from 2000 to 2017 was measured. After cross-sectional dependence and unit root test, the result shows that government health expenditure has no significant impact on reducing child mortality in African countries at all income level.
However, the impact on maternal mortality is only in low income and lower middle income countries.
Furthermore, private health expenditure has significant impact on reducing child and maternal mortality at all income levels.
This study therefore recommends that government at all income level in Africa, especially low income and lower middle income countries, should provide quality health care services and regular immunization for women and children in order to reduce maternal and child mortality in Africa region.