Impact of Government Health Expenditure on Maternal Mortality: An Appraisal of South-Asian-Economies

Abstract

Maternal and child healthcare is a crucial dimension in healthcare delivery. A significant portion of government health expenditure is kept for improvement in infrastructure, service delivery and financial protection to reduce mortality and improve maternal healthcare. The current study aims at assessing the relationship between government health expenditure and maternal mortality in the South Asian context while considering expected years of schooling, fertility rate, government effectiveness, and gross domestic product as control variables. The retrospective panel data has been used from 2002 to 2020 for 8 South Asian Economies. Fixed-effect-model, Pooled-ordinary-least-square, Random-effect-model, Hausman-Test, Fully-Modified-Least-Square and Dynamic-Ordinary-Least-Square techniques have been applied for estimation. The results reveal that government health expenditure is negatively related to maternal mortality across all applied analytical models. The study concludes that a rise in government healthcare expenditure in South Asia can reduce maternal mortality.

Keywords

expected year of schooling fertility rate government effectiveness government health expenditure gross domestic product maternal mortality ratio South Asian economies

Highlights

● The present study highlights the relationship between government health expenditure and maternal mortality ratio in South Asia.

● The study uses fixed effect model and the robustness has been checked using FMOLS and DOLS.

● The study revealed that in South Asia, government health expenditure and government effectiveness reduces maternal mortality. This implies that increased government health expenditure coupled with effective governance in South Asian economies may improve healthcare systems.

● The study also revealed that a rise in fertility rate increases maternal mortality rates in South Asian economies. This may be attributed to the fact that these economies are undergoing demographic transition.

● Fertility rates in these economies are steadily dropping but are still higher when compared to maternal mortality rates.

Introduction

Maternal mortality is one of the most critical and serious challenges. Maternal deaths remain a global concern despite significant progress in medical research and public health efforts. These deaths are a reflection of systemic deficiencies as well as inadequacies in healthcare delivery. The importance of public healthcare expenditure, resource distribution, and their effect on maternal health outcomes are crucial aspects of this complex problem.

Globally, Maternal Mortality (MM) has been reduced, however, in low and middle-income nations there are still major variations carrying the majority of burden. In this regard, government funding for healthcare becomes an important consideration in determining the health of a woman.

Maternal and infant death rates determine the nation’s overall health status. This is an important reflector of the nation’s socioeconomic status, standard of care, population health, and deprivation. Owusu et al,¹ carried out an analysis between the period 2000 to 2017 in which it was revealed that the maternal mortality ratio has reduced between 342 and 311. However, near about 85% of all maternal deaths have been recorded in South-Asian economies and Sub-Saharan economies, which is, higher than other nations worldwide.

Sustainable Development Goal Third (SDG 3), aims to decrease mortality to 70 for every 100 000 live births and to ensure that all people, regardless of age, live healthy lives by 2030.² Thus, a better healthcare system is need of the hour, where significant focus has to be given on the healthcare financing aspect also.^1,3 With investment in the health sector, positive health outcomes are expected, plus improved social and political stability, which in turn creates job opportunities and contributes to the expansion and advancement of the economy.^4
-6

Resource scarcity compels governments in developing nations, especially those in South Asia, to underspend on health in their budget plans.⁷ In recent decades, scientists and researchers have examined the association between spending on health and health outcomes, given the significance of public medical care and its impact on the country’s economy.^8,9 In this regard, numerous investigations have been carried out in the past.^10
-12 Some research studies^13,14 found no such association, while others¹⁵ observed that spending on healthcare enhances health results. Thus, the link between spending on healthcare and health outcomes is unclear because healthcare spending and patterns fluctuate a lot. Therefore, the discussion regarding healthcare spending and health results is still debatable,¹⁶ leading to the need for further investigation. In spite, of the increase in health spending in developing nations, MM remains higher. This has inspired researchers to determine the fundamental cause of maternal deaths to allocate and use health spending wisely.

Maternal mortality can be reduced by ensuring that all birth-giving women should have advanced facilities including skilled healthcare workers and emergency services. Besides, there is a need to improve prenatal care so that it will reduce high-risk pregnancies and train healthcare workers. Furthermore, the engagement of communities in support of birth-giving women and the provision of transportation facilities in times of emergencies is essential to reduce the maternal mortality ratio.

Considering the need for more clarity, the proposed study is a relational analysis of Maternal Mortality (MM) and Government Health Expenditure (GHE) in the South Asian context from 2002 to 2020 by using econometric models like Pooled Ordinary Least Square (POLS), Fixed Effect Model (FEM), Random Effect Model (REM), Hausman Test (HT), Fully Modified Least Square (FMOLS), and Dynamic Ordinary Least Square (DOLS) to observe the long-run effect of healthcare expenditure upon maternal mortality by considering Expected Years of Schooling (EYS), Fertility Rate (FR), Government Effectiveness (GE) and Gross Domestic Product (GDP) as control variables.

Review of Literature

As per the World Health Organization, a woman’s death during or within 42 days of pregnancy due to any pregnancy causes / complications or its management, except death caused by accidents or incidental factors, is termed Maternal Mortality.¹⁷ Till 2020, per day nearly 800 women lost life from pregnancy and childbirth-related avoidable causes. Every 2 min a woman lost her life in 2020 due to pregnancy-related issues. From 2000 to 2020, the global MM was reduced by almost 34%. The literature provides us with a better understanding of how maternal mortality rates have changed over the period in different regions. Huang et al,¹⁸ used the Logistic Regression Model technique for the 2000 to 2019 time period and found that in the United Nations, maternal mortality overall increased especially in American Indian and Black women. Wang et al,¹⁹ find that the US has experienced a change in the underlying causes of pregnancy-related deaths, there has been a rise in mortality resulting from a fall in the cardiovascular system and pre-existing medical disorders. Ward et al,²⁰ studied 200 countries from the 2016 to 2020 data period by using the Global Maternal Health Micro simulation model and found that globally maternal fatalities are anticipated to decrease by 167 in 2023 with 58 nations ranking higher than 140. There has been a reduction in Maternal mortality in other regions as well. Like, in Eastern Europe and Southern Asia, there has been a decline in MM ratio by 70% and 67% respectively between 2000 and 2020.

The government is spending more on health to reduce maternal mortality and other health-related issues. It becomes important to study whether government healthcare spending impacts controlling maternal mortality or not. The association between GHE and MM has been the subject of several studies conducted over some time, either for a single country (time series) or a group of nations (eg, Sub-Saharan Africa, OECD). Although multiple data sets and methodologies were employed in this field, however, the results varied, particularly when it came to the maternal mortality ratio.¹ Numerous relational studies have been conducted analyzing GHE and MM. There is a general assumption that more government health spending can reduce maternal mortality, as the government spends to build good health infrastructure, provide affordable healthcare services, etc. Numerous studies have been conducted to find the relationship between government health spending and maternal mortality. The literature reveals mixed observations for the relationship between the said variables. Sáenz et al²¹ found a reduction in MMR in the region (Latin America and the Caribbean) under analysis in a heterogeneous manner between 1990 and 2019, whereas, the GHE, Government Domestic Product (GDP), and Human Resource for Health (HRH) availability increased heterogeneously. Two-step least squares (2SLS) applied to the demand function for health was employed by Boundioa and Thiombiano²² for 7 nations in the West African Economic and Monetary Union (WAEMU) between 1996 and 2018, wherein, public healthcare expenditure showed a negative significant relationship with MM. Atuhaire et al,²³ examined data period from 1985 to 2019 for Uganda by using the Auto-regressive Distributive lag model and observed that maternal mortality is greatly decreased by domestic government spending on maternal health issues. Chidinma²⁴ in Zimbabwe examined the link between government spending and MM from 1980 to 2011. The findings show that government spending has a statistically significant relationship with maternal death, with higher rates of maternal death being linked with lower government expenditure on health. Another study by Owusu et al,¹ used data from 2000 to 2015 for 177 countries in which the researchers found that there is a drawback of health spending on death across all percentiles. The percentage of infant death falls from 0.19% to 1.45%, and maternal death falls from 0.09% to 1.91%. According to this study, health spending is expected to cut maternal and new-born death rates in lower- and middle-income nations, helping to achieve SDG 3 of guaranteeing good health and well-being for all individuals. Sultana et al,²⁵ employed Vector Autoregression with Exogenous Variables technique to analyze data from Bangladesh, spanning the years 1990 to 2019. The study indicates that there is a positively significant effect of health spending (per capita) and the number of doctors on life expectancy, and child and maternal health. Similarly, Arthur and Oaikhenan²⁶ showed that spending on health significantly increased life expectancy and decreased mortality among children under 5 and mothers. Samuel and Amadi²⁷ analyzed data for Nigeria from 1981 to 2022 and found that the distribution of government funds to the health sector has a negatively significant effect on the rate of death for women. Chidinma²⁴ studied data from 2002 to 2021 for Nigeria by using the Auto Regressive Distributed Lag (ARDL) and Bound testing technique and revealed that the overall government spending on health during the long term has a negative and significant association with MM. In Nigeria (25 States) from 2003 to 2015, Nwankwo²⁸ used Panel Data regression analysis that shows spending on public health is vital for lowering the rate of death of women in Nigeria. Few studies are revealing contrasting results against the above states. Using the regression approach, Patrick et al,²⁹ investigated data for Rwanda from 1990 to 2015 and found a significantly positive relationship between maternal mortality and healthcare expenditure. Aziz et al,³⁰ studied data from 2000 to 2017 for South Asian Economies and observed that a 1% increase in health spending increased MM by 1.95%. Some studies observed no relationship among the variables under discussion.^14,31 Nicholas et al,³² found that there is no discernible effect of health spending on the death of a woman.

From the above studies, it has been found that except for a few studies most of the cited literature showed a negative relationship between GHE and MM. Furthermore, it has been found that different econometric methodologies have been used but not revealing entirely unambiguous results. Thus, to find the latest explicit relationship between Government healthcare spending (GHE) and MM, the study analyzed data from South Asian Economies.

After evaluating the relevant literature and co-occurrence map, the study has also considered the control variables like fertility rate, expected year of schooling, government effectiveness, and gross domestic product. Government healthcare expenditure is one of the important variables and it can directly as well as indirectly impact the maternal mortality of South Asian countries. Thus, it becomes important to study whether government healthcare expenditure positively or negatively impacts maternal mortality in the region under analysis. Therefore, the following hypothesis is proposed and will be tested empirically:

H₀: There is no significant relationship between government healthcare expenditure and maternal mortality

H₁: There is a significant relationship between government healthcare expenditure and maternal mortality

After reviewing the relevant literature there are certain gaps related to the data time period, econometric tools applied, region of study taken, variables, and findings. The Data period taken by the studies Aziz et al,³⁰ Rana et al,³³ Maruthappu et al,³⁴ Huang et al,¹⁸ Boundioa and Thiombiano,²² and Sultana et al,²⁵ is only upto the year 2019 whereas our study made an addition by taking data upto 2020. Most of the studies Maruthappu et al,³⁴ Huang et al,¹⁸ Boundioa and Thiombiano,²² and Sultana et al,²⁵ have used econometric tools like multivariate regression analysis, Logistic regression models; 2-step least squares (2SLS); Vector Autoregression with Exogenous Variables (VARX) model), whereas our studies use FMOLS, DOLS, fixed and random effect model and Hausman Test for robust results. Similarly, studies like Aziz et al,³⁰ Rana et al,³³ Mouinou and Mendy³⁵; Maruthappu et al,³⁴ Huang et al,¹⁸ Chidinma,²⁴ and Boundioa and Thiombiano²² have used the different region for studying such relationship. There is a dearth of studying such relationships in developing regions, particularly in the South-Asian context. Our study contributes by analyzing the relationship in the said required regional context. While studying the relationship between healthcare expenditure and maternal mortality, various other variables have been considered like; economic growth, population sanitation, clean fuel technology, life expectancy, infrastructure, skilled health professionals, incidence of HIV/AIDS, No. of doctors, etc. keeping in view the requirement of controlling other important factors like impact of education, fertility, and effectiveness of the government in terms of its institutions and system, our study contributes the existing literature by taking expected years of schooling, fertility rate, government effectiveness and gross domestic product as control variables. After reviewing the literature different findings have been revealed by various studies in the South-Asian context. Some concluded health expenditure to have a negative relationship with maternal mortality, and some found a positive relationship, whereas, few put forth mixed findings. In such a scenario, there is a need to further evaluate such a relationship, and the current study is an attempt to fill this required gap.

Data and Methodology

Data Source

The variables under analysis have been included in Table 1 under the variable list. The data for each indicator has been acquired from the United Development Program (UNDP), World-Governance-Indicators (WGI), and World-Development-Indicator (WDI). For comprehensive study and evaluation, we have provided both descriptive statistics and an exhaustive list of variables. The research used data spanning from 2002 to 2020 for South Asian Economies (Afghanistan, Bangladesh, Bhutan, India, Maldives, Nepal, Pakistan, Sri Lanka). To understand the relationship of the variables in a larger context, that is why the South Asian region has been selected for analysis, where a good number of countries are still developing and need to improve a larger portion.

Table 1.

Variables and Their Description.

Variable name	Symbols	Indicators’ name	Sources	Expected sign (+,−)
Maternal mortality ratio	lnMMR	Maternal mortality ratio (modeled estimate, per 100 000 live births)	WDI	(NA)
Domestic general government health expenditure	DGHE or GHE	Measured as (% of GDP)	WDI	(+/−)
Expected year of schooling	EYS	For children of school entering the age	(UNDP/ UNESCO)	(−)
Fertility rate	FR	Fertility rate, total (births per woman)	WDI	(+/−)
Government effectiveness	GE	Government Effectiveness: Estimate ( −2.5 to 2.5)	WGI	(−)
GDP per capita growth	GDP	GDP per capita growth (annual %)	WDI	(−)

Source. Authors’ compilation. Variables (MMR) are in natural logarithmic form with (ln) except (DGHE, EYE, FR, GE, GDP).

Techniques Employed for the Data Analysis

The study made use of Panel Ordinary Least Square,²⁷ the Fixed Effect Model,²¹ the Random Effect Model, the Hausman Test, etc. (Boachie et al,³⁶ Kiross et al,³⁷ Boundioa and Thiombiano,²² and Dieleman and Templin³⁸). The reason for using this technique was the non-stationarity behavior of one of the variables specified in the model. Further, the test of robustness was conducted using FMOLS and DOLS.³⁰

Specifications of the Model

It is an important and well-acknowledged fact that an improved understanding of some economic factors influences the formulation of efficient economic policies. This includes a nexus among maternal mortality, domestic government health expenditure, expected year of schooling, fertility rate, government effectiveness, and gross domestic product. The following technique was created to find the link between GHE and MM in South Asian regions over the long term, accounting for factors such as expected year of schooling, fertility rate, government effectiveness, and gross domestic product.

The functional form used to assess how government health spending affects maternal mortality in South Asian Economies is as follows:

MMR = f (DGHE, EYS, FR, GE, GDP)

(1)

To examine how certain below-mentioned variables affect maternal mortality ratio, the variables are converted logarithmically to minimize data fluctuation.

\begin{matrix} l n M M R_{i t} = α_{0} + β_{1} D G H E_{i t} + β_{2} E Y S_{i t} + β_{3} F R_{i t} \\ + β_{4} G E_{i t} + β_{5} G D P_{i t} + μ_{i t} \end{matrix}

(2)

Where MMR, DGHE, EYS, FR, GE, and GDP signify the maternal mortality ratio, expected year of schooling, fertility rate, government effectiveness and gross domestic product respectively. MMR is the dependent variable in this model; the coefficients of the lag in the MMR are β_k (k = 1, 2, 3, 4,5), expected year of schooling, fertility rate, government effectiveness, gross domestic product and µ_it shows the error term. To prevent overfitting and enable the association between our core variables of interest to emerge, our model was limited to these variables.

Pesaran’s CADF test was employed, which has been formulated as:

C I P S (N, T) = N^{- 1} \sum_{i = 1}^{N} t_{i} (N, T)

(3)

where t_i (N, T) represents CD-ADF statistic.

Pedroni³⁹ tests were used for panel cointegration analysis, which is expressed as follows:

y_{i t} = α_{i} + δ_{i t} t + x_{i t} β_{i} + ε_{i t}

(4)

variables $x_{i t}$ and $y_{i t}$ having the dimensions (N*T) ×m and (N*T) ×1, respectively.

Pedroni’s⁴⁰ regression theory serves as the foundation for the models that are suggested for the investigation.

\begin{matrix} l n M M R_{i t} = α_{i} + β_{1} D G H E_{i t} + β_{2} E Y S_{i t} + β_{3} F R_{i t} \\ + β_{4} G E_{i t} + β_{5} G D P_{i t} + μ_{i t} \end{matrix}

(5)

i = 1, 2, . . ., N t = 1, 2, . . ., T

The determinants in this case are cointegrated with slope →β_i, which may or may not be uniform throughout→ i.

where lnMMR_it →maternal mortality ratio with log natural, DGHE _it →domestic general government health expenditure, EYS _it →expected year of schooling, FR _it →fertility rate, GE _it →government effectiveness, GDP_it →gross domestic product with time→ t, and ith →nations. The MMR and its determinants are cointegrated with slopes →βi, which may or may not be uniform throughout→ i.

Pooled ordinary least square (POLS)²⁷ or panel least square are as follows:

y_{i t} = α_{i} + β_{i} x_{i t} + ε_{i t}

(6)

\begin{matrix} l n M M R_{i t} = α_{i} + β_{1} D G H E_{i t} + β_{2} E Y S_{i t} \\ + β_{3} F R_{i t} + β_{4} G E_{i t} + β_{5} G D P_{i t} + μ_{i t} \end{matrix}

(7)

Fixed effect model (FEM) equations are as follows:

y_{i t} = α + β_{i} X_{i t} + u_{i t}

(8)

The random effect model (REM) can be written as follows:

y_{i t} = α + x_{i t} β_{i} + δ_{i t} + μ_{i t}

(9)

$μ_{i t} =$ randomness (no trend); $δ_{i t}$ is different for all cross sections to get individual intercepts with α. It varies sectionally but constantly over time.

Hausman Test

Several econometrics specification difficulties can be addressed with the generic test of Hausman’s⁴¹ specification. Nonetheless, the application that quantifies the individual effects in the panel is the one that receives the most responses from him.

Hypothesis

H₀: Random effect model is better than the fixed effect model

H₁: Fixed effect model is better than the random effect model

When the null hypothesis (H0) is rejected, the FEM Sáenz et al²¹ is used. Otherwise, the REM will be more relevant in this case. The FEM is used first, followed by the REM. And finally, the HT will be applied to opt for the appropriate model either fixed or random.

Fully Modified Ordinary Least Square and Dynamic Ordinary Least Square

These methods were used to check the robustness of our data.³⁰ The Fully Modified Ordinary Least Square (FMOLS) was created by Phillips and Hansen⁴² to provide an ideal co-integrating regression estimate. Due to its ability to account for serial correlation and endogeneity bias, the Pedroni⁴³ for the panel cointegration regressions, the heterogeneous FMOLS estimator was employed in this investigation. For the panel that contains heterogeneous cointegration, FMOLS is the technique that Hamit-Haggar⁴⁴ believes is appropriate. Recalling that the coefficient β of model 1 had a panel FMOLS estimator of:

β_{N T}^{*} - β = (\sum_{i = 1}^{N} L_{22 i}^{- 2} \sum_{i = 1}^{T} {(x_{i t} - {\bar{x}}_{i t})}^{2}) \sum_{i = 1}^{N} L_{11 i}^{- 1} L_{22 i}^{- 1} (\sum_{i = 1}^{T} (x_{i t} - {\bar{x}}_{i t}) μ_{i t}^{*} - T \hat{γ_{i}})

(10)

where,

μ_{i t}^{*} = μ_{i t}^{*} - \frac{\hat{L_{21 i}}}{\hat{L_{21 i}}} △ x_{i t}, \hat{γ_{i}} = \hat{τ_{21 i}} \hat{Ω_{21 i}^{0}} - \frac{\hat{L_{21 i}}}{\hat{L_{21 i}}} (τ_{22 i} + \hat{Ω_{21 i}^{0}}),

(11)

and ${\hat{L}}_{i}$ →was the lower triangulation of $\to \hat{Ω_{i}}$ . The asymptotic distribution of the panel FMOLS estimate obtained by Pedroni and the Dynamic OLS estimator were identical. To verify that the results are consistent, the estimations for DOLS and FMOLS^30,45 were carried out as demonstrated.

Result and Discussion

Empirical Results and Analysis

Descriptive Statistics

Table 2 lists the variables utilized in this investigation along with their descriptive statistics. The results for the study nations highlight the max. & min. values of MM are approximately 7.152 and 3.367, respectively. The health expenditure of the examined countries ranged from a min. of 0.084 to a max. of 9.077, with a standard deviation of 1.565 and a mean value of around 1.599. The mean value of the expected year of schooling was valued at 10.702 with a standard deviation of around 2.253. The mean values for fertility rate, government effectiveness and gross domestic product were 3.043, −0.436 & 3.588 resp., with standard deviations (S.D) of 1.363, 0.583, and 5.498.

Table 2.

Table for descriptive statistics.

Variables	N	Mean	SD	Min.	Max.
LMMR	152	5.131	0.979	3.367	7.152
DGHE	152	1.599	1.565	0.084	9.077
EYS	152	10.702	2.253	5.278	14.140
FR	152	3.043	1.363	1.433	7.339
GE	152	−0.436	0.583	−1.609	0.835
GDP	152	3.588	5.498	−34.831	22.020

Source. Estimations by the Author’s.

Panel Data Results

Cross-Sectional Dependency Estimates and Unit Root Tests

The Cross-Sectional Dependency (CSD) test was used in this investigation before cointegration verification. Some have suggested that in data series with cross-section dependency, traditional unit root tests do not provide accurate results. Since all of the variables’ findings were shown to be significant at the 1% significance level, it was evident from the cross-sectional dependency test results that there was cross-sectional dependency among the variables (see Table 3). It recommends using a cross-sectional unit-root-test (CIPS and CADF) rather than the traditional Liven, Lin, and Chu (LLC); Im, Pesaran, and Shin (IPS) unit root test for panel data with cross-section dependency. As a result, the current study made use of the CIPS and CADF results, which demonstrated that every variable became stable at the first difference and assumed cross-section dependency within the panel data.^46,47

Table 3.

Cross Sectional Dependency Test and Second Generation (CIPS, CADF) Unit Root Test Results.

Variables	CD tests		Second-generation unit root tests
	B-P LM	P-S LM	CIPS test		CADF test
	t-Stat	t-Stat	Level	First diff. Δ	Level	First diff. Δ
LMMR	477.37^***	60.05^***	−1.747	4.384^***	−1.520	−2.718^***
DGHE	122.94^***	12.69^***	−2.785^***	-	−2.785^***	-
EYS	465.30^***	58.44^***	−2.090	−3.584^***	−2.090	−3.584^***
FR	75.662^***	6.370^***	−1.578	−2.874 ^***	−1.548	−2.607^***
GE	461.61^***	57.95^***	−2.279^**	-	−2.279^***	-
GDP	104.80^***	10.26^***	−3.45^**	-	−3.58^***	-

Source. Estimation by Author’s (***→1%, **→5% and *→10% indicates significance levels).

Panel Cointegration Test

Panel data cointegration was valid to use when all variables’ stationarity properties were established. The Pedroni panel cointegration test was employed in the study to verify the cointegration.⁴⁸ Six of the 11 tests were significant at the 1%, and 5%, levels of significance with intercept, and with intercept and trend, according to Table 4’s results. These test findings clearly show that variables are co-integrated over the long term.

Table 4.

Results of Panel Cointegration Tests.

Testing Pedroni’s residual cointegration
	With intercept		With trend & intercept
	Stat	P-value	Stat	P-value
Within-dimension
Panel v-Statistic	−0.351	.637	0.001	.499
Panel rho-Statistic	1.524	.936	2.113	.982
Panel PP-Statistic	−1.867^**	.030	−3.827^***	.000
Panel ADF-Statistic	−1.982^**	.023	−4.099^***	.000
Weighted
Panel v-Statistic	−0.826	.795	−1.179	.880
Panel rho-Statistic	1.318	.906	1.912	.972
Panel PP-Statistic	−2.046^**	.020	−3.125^***	.000
Panel ADF-Statistic	−2.110^***	.017	−3.244^***	.000
Between-dimension
Group rho-Statistic	2.540	.994	3.103	.999
Group PP-Statistic	−2.197^***	.014	−4.281^***	.000
Group ADF-Statistic	−2.016^**	.021	−3.201^***	.000

Source. Estimation by Author’s (***→1% and **→5% indicates significance levels).

Panel Estimation Results

Fixed Effect and Random Effect Results

Panel regression results are shown in Table 5. The table displays 3 sets of estimates: (1) POLS estimates; (2) FEM estimates; and (3) REM estimates.

Table 5.

Estimates of Parameters for Panel Regression Model.

Variables	Panel OLS	P-value	Fixed effect model	P-value	Random effect model	P-value
DGHE	−0.206^***	.000	−0.053^***	.008	−0.106^***	.000
EYS	0.097^***	.000	−0.171^***	.000	−0.033^***	.000
FR	0.397^***	.000	0.205^***	.000	0.436^***	.000
GE	−0.386^***	.000	−0.116*	.095	−0.323^***	.000
GDP	0.003	.683	−0.005^**	.022	−0.005^***	.011
Constant	3.031^***	.000	6.394^***	.000	6.134^***	.000
R-squared	.759 92.109 .000		.984 732.815 .000		.692 65.655 .000
F-statistic
Prob(F-statistic)

Source. Estimation by Author’s (***→1%, **→5% and *→10% indicates significance levels).

The results for the POLS reveal that GHE and GE show a negative effect on MM while EYS FR and GDP show a positive effect on MM. Thus, an increase in government health expenditure and government effectiveness reduces maternal mortality. This implies that increased government health expenditure coupled with effective governance in South Asian economies may improve healthcare systems and result in improved access to quality maternal health services and thus may lower maternal mortality rates in the study area. In contrast, the rise in fertility rate increases maternal mortality rates in South Asian economies. This may be attributed to the fact that these economies are undergoing demographic transition. Fertility rates in these economies are steadily dropping but are still higher when compared to maternal mortality rates. When there is an increase in EYS and GDP the maternal mortality rate also increases as indicated by POLS. However, as indicated by the Fixed effect model there is a negative impact of EYS and GDP on maternal mortality rate. Generally, with higher education levels the maternal mortality rates must be reduced due to better healthcare awareness and improved decision making. The fixed effect model controls for country-specific factors so they may better point out the positive impact of education on maternal health. The study for fixed effect reveals that government health expenditure, expected year of schooling, government effectiveness and gross domestic product have negatively influenced maternal mortality in South Asian economies by 0.053%, 0.171%, 0.116%, and 0.005% respectively, and similar results are also revealed by random effects, with coefficients 0.106%, 0.033%, 0.323%, and 0.005% respectively. Thus, the test results reveal that an increase in health spending, expected year of schooling, government effectiveness and gross domestic product reduces maternal mortality while in contrast, the fertility rate shows unexpected results and reveals that an increase in fertility rate increases maternal mortality.^33,49
-51 The REM indicates a similar relationship in sample nations.

Hausman’s Test Results

As shown in Table 6, the result of the HT statistic (Hausman⁴¹): A low P-value favors the FEM and works against the null hypothesis that the REM is consistent. The p-value is lower than .05 shows that among tested models the FEM is appropriate in this case. The R² coefficient is used in determining the explanatory power of our independent variables as related to changes in the dependent variable. For our model, R² is .98 under the FEM. This means that about 98% of the variation in MM is explained by changes in government health spending, expected year of schooling, fertility rate, government effectiveness, and gross domestic product variables. Nevertheless, because the nations are different and do not have the same demographics, healthcare systems, medical advancements, or disease prevalence, heterogeneity and cross-sectional dependence affected the panel data. Outcomes based on fixed effects could be biased. FMOLS and DOLS were thus used in the investigation. Autocorrelation and heteroscedasticity in the panel data can be effectively overcome with these techniques.

Table 6.

Hausman’s Test.

Test summary		Chi-sq. statistic	Chi-sq. d.f.	Prob.
Cross-section random		485.17	5	.0000^***
Variable	Fixed	Random	Var (Diff.)	Prob
DGHE	−0.0523	−0.1063	0.0000	.1381
EYS	−0.1714	−0.0334	0.0000	.0036
FR	0.2054	0.4360	0.0001	.0095
GE	−0.1158	−0.3231	0.0003	.0402
GDP	−0.0051	−0.00537	0.0000	.8291

Source. Estimation by Author’s (***→1% indicates significance level).

Fully Modified Ordinary Least Square (FMOLS) and Dynamic Ordinary Least Square (DOLS) Results

The estimates for FMOLS and DOLS are reported in Table 7, which shows that while the coefficients’ magnitudes vary, their signs remain identical. It was anticipated that a 1% increase in health spending would have a negative relationship of 0.07% of the MM on FMOLS and 0.06% of the MM on DOLS estimates for economies in South Asia. This implies that For FMOLS and DOLS, respectively, a 1% increase in fertility rate resulted in a 0.19% and 0.09% increase in maternal mortality, according to the fertility rate statistics. Results were found to be negatively significant when compared to the expected year of schooling, government effectiveness and gross domestic product as 0.17%, 0.17%, and 0.008% for FMOLS and 0.17%, 0.12% and 0.009% for the DOLS estimator, respectively.

Table 7.

Results for Long-Run FMOLS and DOLS.

Variables	FMOLS				DOLS
Variables	Coefficient	Std. error	T-stat	P-value	Coefficient	Std. error	T-stat	P-value
DGHE	−0.0705^**	0.0294	−2.3982	.0179	−0.0667^***	0.0250	−2.6663	.0091
EYS	−0.1752^***	0.0240	−7.2847	.0000	−0.1792^***	0.0234	−7.6428	.0000
FR	0.1999^***	0.0642	3.1704	.0019	0.0940	0.0747	1.2583	.2115
GE	−0.1766	0.1097	−1.6101	.1098	−0.1289	0.0969	−1.3308	.1866
GDP	−0.0082^**	0.0033	−2.4312	.0164	−0.0091^**	0.0038	−2.3985	.0185

Source. Estimation by Author’s (***→1%, **→5% and *→10% indicates significance levels).

The impact government health expenditure variables on the maternal mortality in long run, can be seen from the finding of the FEM, REM, FMOLS, DOLS model presented in Tables 5 and 7 and summarized in Table 8. As evident from Table 8, which depicts the impact of the independent variable of the study on maternal mortality, DGHE, EYS, GE, and GDP have a negative long-term effect on maternal mortality. On the other hand, FR have a positive effect on maternal mortality in the long run.

Table 8.

Summary of Long Run Impact of Independent Variables on the Dependent Variables.

Independent variable	Dependent variable (long run)
Independent variable	FEM & REM	FMOLS & DOLS
DGHE	−	−
EYS	−	−
FR	+	+
GE	−	−
GDP	−	−

Note. Summarized based on finding of the FEM, REM, FMOLS, DOLS model from Tables 5 and 7.

“+” and “−” signs indicate the positive and negative impact, respectively.

Dependent variable: Maternal Mortality (lnMM).

Discussion

The present study highlights the relationship between MM and GHE.³⁰ We used the data for 8 South Asian Economies time spanning from 2002 to 2020 for FEM,²¹ REM, FMOLS, and DOLS³⁰ estimation techniques.^22,36
-38 The results showed that there is a fall in MM by increase in government health spending. Results from POLS²⁷ analysis showed that expenditure on health and government effectiveness are important factors in reducing MM for South Asian Economies. This implies that increased government health expenditure coupled with effective governance⁵¹ in South Asian economies improves healthcare systems which results in improved access to quality maternal health services⁵² and thereby significantly reduces maternal mortality rates in the study area. Fertility rate and MM have been found positively related. Here the relationship found is not to our surprise, since both the variables show decreasing values.⁵³ Theoretically, both variables should be positively related, because with declining fertility, the possibility of risky pregnancies also declines.^49,50 PLOS results reveal a contrasting picture of a positive relationship with MM for the controlling variable EYS and GDP.

To get more reliable results, long-run FMOLS and DOLS analysis was carried out, where in addition to the above-discussed results, EYS and GDP is found to be negatively related to MM. More years of schooling is expected to reduce MM on the assumptions of increasing knowledge related to better and necessary healthcare utilization,³³ together with awareness of maternal healthcare facilities.⁵⁴ Further assumptions of more education and more financial health^55,56 and better access to resources⁵⁷ will lead to reduced MM.

Conclusion and Policy Implications

Our study has established that government health expenditure lowers the mortality rate in the study area. This indicates that the existing expenditure in South Asian economies is effective. Therefore, in these economies, the government may consider to increase the proportion of their health budget. This additional investment may increase the benefits of mortality reduction and may involve investment toward the development of quality healthcare infrastructure. Besides, it can also contribute toward increasing accessibility to basic healthcare facilities and quality services and prioritizing preventive healthcare measures. Thus, the effective reduction of mortality rates in South Asia through government health expenditure highlights the dynamic role that healthcare investment plays in fostering the overall well-being of the economy. The study further revealed that the government’s effectiveness reduces the maternal mortality ratio in South Asia. This indicates that through effective governance, the Government in these economies prioritizes healthcare investment and their policies are designed toward increasing access to maternal healthcare services. Thus, in South Asia effective governance needs to be sustained for better health outcomes in general and reduced maternal mortality rate in particular. In addition to the above, the study also revealed that education plays an important role in reducing maternal mortality rates, so policymakers in this region need to design policies focusing on more female-inclusive-education while controlling female dropout ratio (particularly increasing number of years of schooling for marginalized population of the region). The study has also pointed out that in this region, fertility rates increase the mortality rates. This indicates that the high fertility rates in this region put pressure on existing healthcare services and therefore leave women in this region to face maternal health risks. Therefore, more investment is required to build adequate healthcare infrastructure to reduce maternal mortality.

Limitations and Future Directions

Since the study has concluded based on analyzed data from 2002 to 2020, due to the limitation of non-availability of previous data related to a few variables for the selected economies, there is an immense scope for more analysis for the same economies by considering large data period. The study can also be extended by including other important relevant control variables that affect maternal mortality rates. The study used the panel data set and future research can be based on longitudinal data, it can be associative research and it can be extended to study the mediate or moderating relationship between the variables.

Supplemental Material

sj-docx-1-inq-10.1177_00469580251339069 – Supplemental material for Impact of Government Health Expenditure on Maternal Mortality: An Appraisal of South-Asian-Economies

Supplemental material, sj-docx-1-inq-10.1177_00469580251339069 for Impact of Government Health Expenditure on Maternal Mortality: An Appraisal of South-Asian-Economies by Azhar Jan, Meenakshi Gupta and Sheikh Aftaab-Ul-Maroof in INQUIRY: The Journal of Health Care Organization, Provision, and Financing

Footnotes

Acknowledgements

We are sincerely thankful to Editor-in-Chief Prof. Umair Shafique and four anonymous reviewers for insightful comments.

ORCID iDs

Azhar Jan

Meenakshi Gupta

Sheikh Aftaab-Ul-Maroof

Author Contributions

Azhar Jan: Conceptualization, Data curation, Validation, Writing—original draft. Dr. Meenakshi Gupta: Conceptualization, Investigation, Supervision, Validation, Writing—review & editing. Dr. Sheikh Aftaabul Maroof: Conceptualization, Validation, Writing—review & editing.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Data Availability Statement

Data shall be available at:

Supplemental Material

Supplemental material for this article is available online.

References

Owusu

Sarkodie

Pedersen

PA.

Relationship between mortality and health care expenditure: sustainable assessment of health care system. PLoS One. 2021;16(2):e0247413.

Le Blanc

. Towards integration at last? The sustainable development goals as a network of targets. Sustain Dev. 2015;23(3):176-187.

Jan

Gupta

Aftaab-Ul-Maroof

A comparative analysis of healthcare funding sources and their contribution to overall health and well-being for sustainable development. Acad Mark Stud J. 2025;29(2):1-11.

World Health Organization (WHO). Global Health Observatory (GHO) Data Infant Mortality Situation And Trends. World Health Organisation. 2017 https://www.who.int/data/gho

Aziz

Impact of COVID-19 on peoples’ willingness to consume wild animals: Empirical insights from China. One Health. 2021;12:100240.

Yao

Zhang

Aziz

Investigating the links between vaccination against COVID-19 and public attitudes toward protective countermeasures: Implications for public health. Front Public Health. 2021;9:702699.

World Health Organization. World Health Statistics. World Health Organization; 2010. Accessed 2023. https://www.who.int/publications/i/item/9789241563987

Jan

Gupta

Aftaab-Ul-Maroof

Panel causality analysis between public health expenditure and determinants of economic growth in South Asian economies. Acad Mark Stud J. 2025;29(1):1-5.

Jan

Parah

Malik

BA.

IEFHAC: Image encryption framework based on hessenberg transform and chaotic theory for smart health. Multimed Tools Appl. 2022;81(13):18829-18853.

10.

Babazono

Hillman

AL.

A comparison of international health outcomes and health care spending. Int J Technol Assess Health Care. 1994;10(3):376-381.

11.

Nair

VM.

Health in South Asia: future of Kerala depends on its willingness to learn from past. BMJ. 2004;328(7454):1497.

12.

Maroof

Sangmi

MUD.

Assessing financial impact of a health intervention program and controlling spillover effects. Decision. 2021;48(1):15-25.

13.

Burnside

Dollar

Aid, the Incentive Regime, and Poverty Reduction. World Bank Publications; 1998 (No. 1937).

14.

Filmer

Pritchett

The impact of public spending on health: does money matter?

Soc Sci Med. 1999;49(10):1309-1323.

15.

Baldacci

Clements

Gupta

Cui

Social spending, human capital, and growth in developing countries. World Dev. 2008;36(8):1317-1341.

16.

Novignon

Olakojo

Nonvignon

The effects of public and private health care expenditure on health status in Sub-Saharan Africa: new evidence from panel data analysis. Health Econ Rev. 2012;2(1):8.

17.

WHO, UNICEF, UNFPA, the World Bankand the United Nations Population Division. Trends in Maternal Mortality: 1990 to 2013. World Health Organization; 2014; 201456.

18.

Huang

Spence

Abenhaim

HA.

Racial disparities in national maternal mortality trends in the United States from 2000 to 2019: a population-based study on 80 million live births. Arch Gynecol Obstet. 2024;309(4):1315-1322.

19.

Wang

Rexrode

Florio

Rich-Edwards

Chavarro

JE.

Maternal mortality in the United States: Trends and opportunities for prevention. Annu Rev Med. 2023;74(1):199-216.

20.

Ward

Atun

King

Sequeira Dmello

Goldie

SJ.

Simulation-based estimates and projections of global, regional and country-level maternal mortality by cause, 1990–2050. Nat Med. 2023;29(5):1253-1261.

21.

Sáenz

Nigenda

Gómez-Duarte

Rojas

Castro

Serván-Mori

Persistent inequities in maternal mortality in Latin America and the Caribbean, 1990–2019. Int J Equity Health. 2024;23(1):96.

22.

Boundioa

Thiombiano

Effect of public health expenditure on maternal mortality ratio in the West African Economic and Monetary Union. BMC Womens Health. 2024;24(1):109.

23.

Atuhaire

Ekirapa-Kiracho

Mutenyo

The effect of healthcare expenditure on maternal mortality in Uganda. Tanzan Econ Rev. 2023;13(2):165-183.

24.

Chidinma

GAC

. Public health expenditure and maternal mortality in Nigeria. Eur J Public Health Stud. 2024;7(1):15-32.

25.

Sultana

Hossain

Khan

Saha

Amin

Haque Prodhan

MM.

Effects of healthcare spending on public health status: an empirical investigation from Bangladesh. Heliyon. 2024;10(1):e24268.

26.

Arthur

Oaikhenan

HE.

The effects of health expenditure on health outcomes in Sub-Saharan Africa (SSA). Afr Dev Rev. 2017;29(3):524-536.

27.

Samuel

Amadi

EI.

Exploring the effectiveness of healthcare expenditure in reducing maternal mortality in Nigeria. Int J Soc Sci Res. 2024;7(04):46-53.

28.

Nwankwo

CE.

The effects of public health spending on maternal mortality in Nigeria. J Econ Sustain. 2018;9(20):2222-2855.

29.

Patrick

Zaman

MSU

Afzal

Mahsud

Hanifatu

MN.

Factors that affect maternal mortality in Rwanda: A Comparative Study with India and Bangladesh. Comput Math Methods Med. 2022;2022:1-9.

30.

Aziz

Sarker

Sui

Exploring the role of health expenditure and maternal mortality in South Asian countries: an approach towards shaping better health policy. Int J Environ Res Public Health. 2021;18(21):11514.

31.

Musgrove

Public and Private Roles in Health: Theory and Financing, Vol. 339. World Bank, HNP Discussion Paper Patterns; 1996.

32.

Nicholas

Edward

Bernardin

The effect of health expenditure on selected maternal and child health outcomes in Sub-Saharan Africa. Int J Soc Econ. 2016;43(12):1386-1399.

33.

Rana

Alam

Gow

Health expenditure, child and maternal mortality nexus: a comparative global analysis. BMC Int Health Hum Rights. 2018;18(1):29.

34.

Maruthappu

KYB

Williams

Atun

Agrawal

Zeltner

The association between government healthcare spending and maternal mortality in the European Union, 1981–2010: a retrospective study. BJOG. 2015;122(9):1216-1224.

35.

Mouinou

MNS

Mendy

Public health expenditure and health outcomes in Comoros: an asymmetric evidence using multiple threshold nonlinear ARDL model. Theor Econ Lett. 2024;14(4):1564-1582.

36.

Boachie

Põlajeva

Frimpong

AO.

Infant mortality in low- and middle-income countries: does government health spending matter?

J Dev Policy Pract. 2020;5(1):54-73.

37.

Kiross

Chojenta

Barker

Loxton

The effects of health expenditure on infant mortality in Sub-Saharan Africa: evidence from panel data analysis. Health Econ Rev. 2020;10(1):5.

38.

Dieleman

Templin

Random-effects, fixed-effects and the within-between specification for clustered data in observational health studies: A simulation study. PLoS One. 2014;9(10):e110257.

39.

Pedroni

Critical values for cointegration tests in heterogeneous panels with multiple regressors. Oxf Bull Econ Stat. 1999;61(S1):653-670.

40.

Pedroni

Fully modified OLS for heterogeneous cointegrated panels. Adv Econ. 2000;15:93-130.

41.

Hausman

JA.

Specification tests in econometrics. Econometrica. 1978;46:1251-1271.

42.

Phillips

PCB

Hansen

BE.

Statistical inference in instrumental variables regression with I (1) processes. Rev Econ Stud. 1990;57(1):99-125.

43.

Pedroni

Fully modified OLS for heterogeneous cointegrated panels. In: Baltagi BH, Fomby TB and Carter Hill R (eds) Nonstationary Panels, Panel Cointegration, and Dynamic Panels. Emerald Group Publishing Limited; 2001;93-130.

44.

Hamit-Haggar

Greenhouse gas emissions, energy consumption and economic growth: A panel cointegration analysis from Canadian industrial sector perspective. Energy Econ. 2012;34(1):358-364.

45.

Dutta

Gupta

Sarkar

Sengupta

PP.

Some determinants of infant mortality rate in SAARC countries: An empirical assessment through panel data analysis. Child Indic Res. 2020;13:2093-2116.

46.

Barbieri

Panel unit root tests under cross-sectional dependence: an overview. J Stat Adv Theory Appl. 2009;1(2):117-158.

47.

Dritsaki

The relationship between health expenditure, CO2 emissions, and economic growth in G7: Evidence from heterogeneous panel data. J Knowledge Econ. 2024;15(1):1-4911.

48.

Pesaran

Shin

Testing for unit roots in heterogeneous panels. J Econom. 2003;115(1):53-74.

49.

Winikoff

Sullivan

Assessing the role of family planning in reducing maternal mortality. Stud Fam Plann. 1987;18(3):128-143.

50.

Stover

Ross

How increased contraceptive use has reduced maternal mortality. Matern Child Health J. 2010;14:687-695.

51.

Ruiz-Cantero

Guijarro-Garvi

Bean

Martínez-Riera

Fernández-Sáez

Governance commitment to reduce maternal mortality. A political determinant beyond the wealth of the countries. Health Place. 2019;57:313-320.

52.

Austin

Langer

Salam

Lassi

Das

Bhutta

ZA.

Approaches to improve the quality of maternal and newborn health care: an overview of the evidence. Reprod Health. 2014;11:1-9.

53.

Jain

AK.

Measuring the effect of fertility decline on the maternal mortality ratio. Stud Fam Plann. 2011;42(4):247-260.

54.

Roy

Khatun

Effect of adolescent female fertility and healthcare spending on maternal and neonatal mortality in low resource setting of south Asia. Health Econ Rev. 2022;12(1):47.

55.

Sandiford

Cassel

Montenegro

Sanchez

The impact of women's literacy on child health and its interaction with access to health services. Popul Stud. 1995;49(1):5-17.

56.

Pinto

Scheer

Tuqa

Ebrahim

Abel

Mukherjee

DS.

Does health intervention ameliorate the effects of poverty related diseases? The role of female literacy. J Trop Pediatr. 1985;31(5):257-262.

57.

Carvalho

Evans

Girls’ Education and Women’s Equality: How to Get More Out of the World’s Most Promising Investment. Center for Global Development; 2022.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.02 MB