Health Expenditure, Institutional Quality, and Under-Five Mortality in Sub-Saharan African Countries

By inergency On Mar 12, 2024

[ad_1]

4.1. Descriptive Statistics

Table 2 shows that the standard deviations are large enough to explore variance in the data for the variables of concern. Table 2 indicates that the average U5MR value between 2000 and 2021 is 90.2 per 1000 births. Table 2 suggests that the public health expenditure (PHE) has a lower (5.2%) average value, while the domestic private health expenditure has the highest average value (48.02%) in SSA countries between 2000 and 2021. Again, descriptive statistics show that health spending is primarily out-of-pocket, with an average of 40.5% between 2000 and 2021. Furthermore, Table 2 demonstrates that 28.24% of health spending in SSA countries is external. The average HIV prevalence and undernourishment rates were 5% and 21.7%, respectively. According to Table 2, 75% of children in SSA nations received an immunization vaccine before reaching one year. Between 2000 and 2021, the average value for secondary education enrolment is 44.3%. The average ICT indices (FBS, ITU, and MCS) are 0.99%, 12.1%, and 49.4/100 people, respectively. The average GDP per capita is $2120. Between 2000 and 2021, the average point for the institutional quality index was −0.004.

Table 2.
Descriptive Statistics.

Variable	Obs	Mean	Std. Dev.	Min	Max
U5MR	1034	90.12186	42.54212	14	229
phe	956	5.224079	2.264309	1.26	20.41
hedp	956	48.01509	18.64855	8.34748	87.93836
heoup	956	40.44704	20.83885	2.993242	84.18211
heext	953	28.24149	32.24391	0	228.0008
phiv	990	5.00798	6.693593	0.1	29.8
pun	720	21.71611	13.81633	3.1	70.9
imm	1034	75.87911	18.74111	19	99
mam	987	502.4762	290.1703	3	1682
senr	620	44.2679	23.2978	5.51	114.71
fbs	743	0.9869448	3.404632	0	38.77
itu	1002	12.09936	16.17075	0.01	81.59
mcs	1026	49.37572	42.79125	0	185.56
ict	725	0.3427037	1.511278	−1.270262	10.10088
gdppc	998	2120.417	2785.312	255.1	16,747.34
inst	986	−0.0039528	2.212385	−6.015036	5.546552

The correlation coefficient matrix (see Appendix A) shows that the under-five mortality rate is negatively correlated with public and external health expenditure while positively correlated with domestic private health and out-of-pocket health expenditure. The U5MR is also negatively correlated with secondary education enrolment, ICT, immunization, GDP per capita, and institutional quality. However, the prevalence of HIV undernourishment and maternal mortality are positively correlated with under-five mortality rates in SSA countries. Again, the variance inflation factor (VIF) values (see Appendix B) of the explanatory variable are less than 5 or 10, which implies the absence of multicollinearity in the analysis of the independent variable components. This also aligns with the recommendation of Gujarati and Porter (2010) [77] and Asteriou and Hall (2016) [56] The correlation coefficient matrix and the VIF test indicate that all the explanatory variables are not linearly dependent. They may thus be used in the same model of regression.

The Im Pesaran and Augmented Dickey–Fuller (ADF) stationarity test results (see Appendix C) indicate that variables such as out-of-pocket health expenditure, prevalence of HIV, prevalence of undernourishment, ICT, and GDP per capita are stationary after the first difference. At the same time, the rest of the variables are stationary at levels. Thus, the paper applied the differencing transformation for the variables that are stationary after first differencing, as recommended by Johnston and DiNardo (1997) [78], Brooks (2008) [79], Greene (2010) [80], and Asteriou and Hall (2016) [56].

Determination tests of the appropriate estimator between DGMM and SGMM were performed on each health expenditure indicator, based on the coefficients of the lagged dependent variables (Bond et al., 2001) [64]. Thus, each model’s coefficients of the lagged dependent variables are summarized in Appendix E. The results (see Appendix E) of the DGMM coefficients are lower than those of fixed effect (FE), indicating that the SGMM is the appropriate and efficient estimation technique for the study.

4.2. Empirical Results

This section discusses the empirical results of the study in Table 3 and Table 4. Table 3 presents empirical results of the direct effect of health expenditure on the U5MR, and Table 4 shows the impact of institutional quality on the U5MR-health expenditure nexus in SSA countries. In this study, the impact of health expenditure is divided into four categories, including public health expenditure (PHE), domestic private health expenditure (HEDP), out-of-pocket health expenditure (HEOUP), and external health expenditure (HEEXT). Model 1 in Table 3 is the benchmark model that shows empirical results of the direct impact of public health expenditure on the U5MR in SSA. Models 2 through 4 show the effects of HEDP, HEOUP, and HEEXT on the U5MR, respectively.

Table 3 indicates that the under-five mortality rate for the previous year increases the under-five mortality rate for the current year by 0.01% at a 1% significance level. This result aligns with Dhrifi (2020) [41], Langnel and Buracom (2020) [26], and Hadipour, Delavari, and Bayati (2023) [40], who argue that infant mortality is persistent in developing economies. This is attributable to limited access to healthcare services and infrastructure and shortages of healthcare facilities and trained personnel, especially in remote areas of SSA.

The results show that public health expenditure and external health expenditure have a negative impact on the U5MR. Thus, a 1% increase in public and external health expenditure is associated with a 0.1% and a 0.03% decrease in the under-five mortality rate at a 10% significance level. This indicates that allocating a reasonable proportion of public health expenditure enhances lower under-five mortality rates in SSA countries. The negative effect of external health on the U5MR is attributable to the fact that external health expenditure can provide additional funding for healthcare infrastructure, services, and programs, improving child survival rates. The results align with findings by [23,24,28,39,40,81].

Contrariwise, the domestic private health expenditure is positively signed. Thus, a percentage increase in HEDP worsens the U5MR by 0.2% at 5% significance levels. However, the out-of-pocket health expenditure has a positive but insignificant effect on the U5MR. Therefore, a positive impact of domestic private health expenditure on the U5MR suggests that personal spending typically serves as a coping mechanism when public healthcare systems are insufficient, compromising child healthcare and exacerbating child mortality. The finding aligns with Kulkarni (2016) [42] and Logarajan et al. (2022) [39]. Out-of-pocket expenditures harm infant mortality by potentially undermining health-seeking behaviors [44]. The positive effect of out-of-pocket health expenditure on the U5MR is attributable to high out-of-pocket healthcare expenses, which can act as a significant financial barrier, preventing many families from seeking timely medical care for their sick children. When parents or caregivers cannot afford necessary healthcare services, children may not receive prompt treatment, leading to worse health outcomes.

Empirical results show that a percentage increase in the prevalence of HIV and maternal mortality increases U5MR by 0.1% and 0.3–0.5%, respectively, at a 1% significance level. The positive effect of HIV on the U5MR accords with Anyanwu and Erhijakpor (2009) [22], Chihana et al. (2013) [69], Fowkes et al. (2016) [82], and Kiross et al. (2021) [27] who suggest that the prevalence of HIV reduces the chances of child survival. Studies such as Akinlo and Sulola (2019) [5] and Chewe and Hangoma (2020) [25] argue that HIV has a negative effect on under-five mortality due to an increase in the provision of antiretroviral during pregnancy, which has reduced the risk of mother-to-child transmission. However, the positive impact on the current empirical research is because of the lack of awareness and high HIV stigma in society and communities. Thus, without proper HIV awareness and growing HIV stigma amongst women, especially young people, young mothers become ignorant and afraid to expose their HIV status while pregnant, which leads to them infecting the unborn child, which may cause infant health complications, resulting in increased under-five mortalities. Again, limited access to medication exacerbates the under-five mortality rates in the region.

Regarding maternal mortality, the child is at risk of losing a primary caregiver when its mother dies. This can jeopardize the child’s access to necessary care, nutrition, and emotional support, all of which are essential for a child’s general development and survival. The positive effect of maternal mortality is consistent with Finlay et al. (2015) [75], Moucheraud et al. (2015) [76], and Scott et al. (2017) [83], who opine that maternal mortality compromises child survival. The results show that the prevalence of undernourishment is positively signed. Thus, a 1% change in undernourishment prevalence is associated with a 0.1% to 0.2% increase in the U5MR. This implies that the prevalence of undernourishment is an under-five mortality-promoting factor in sub-Saharan countries. The result aligns with Ssozi and Amlani (2015) [44] and Djoumessi (2022) [84], who suggest that undernourishment aggravates the mortality rate.

The results show that immunization has an asymmetric effect on under-five mortality in sub-Saharan countries. A 1% increase in immunization reduces the U5MR by 0.2% at a 10% significance level. The negative impact of immunization accords with Arthur and Oaikhenan (2017) [43], Akinlo and Sulola (2019) [5], Dhrifi (2020) [41], and Ayipe and Tanko (2023) [28], who posit that immunizations and other public health services shield children from harmful illnesses like polio, diphtheria, measles, and tetanus. Thus, immunization of children against these diseases lowers their risk of contracting and developing infections, which reduces the under-five mortality rate in SSA. A positive effect of immunization on under-five mortality could be attributed to low immunization and inequities in vaccination coverage [85] rates within the SSA population.

Again, a percentage increase in education reduces the U5MR by 0.1–2% at a 1% significance level. This implies that improved education for women makes women more knowledgeable about maternal health, including safe delivery techniques, appropriate nutrition throughout pregnancy, and the significance of prenatal and postnatal care. The result aligns with Grossman’s (1972, 2000) [29,30] theory, which postulated that education positively affects health demand. A negative impact of education on under-five mortality aligns with Ouedraogo et al (2020) [45], Owusu et al. (2021) [38], Ouedraogo, Simon and Kiragu (2022) [86] and Moradhvaj and Samir (2023) [87], who documented evidence of the negative impact of education on infant mortality. Once more, postponing parenthood can lower the chance of risks during pregnancy and delivery, which can help to improve results for the health of both mother and child.

The result indicates that ICT is an U5MR-reducing factor in SSA countries. Thus, a 1% increase in ICT reduces the U5MR by 0.12–0.4% at a 5% significance level. This implies that increased ICT can support telemedicine services, enabling medical professionals to diagnose and treat patients, including children, from a distance. A health information system that tracks and monitors child health indicators assists in identifying health trends and better allocating resources, ensuring that healthcare interventions are provided to the children most in need. Additionally, ICT can offer pregnant mothers access to training, as well as educational materials to nurses and community health workers, improving health outcomes and reducing under-five mortality rates. This result accords with Dutta et al. (2019) [50], Kouton, Bétila, and Lawin (2021) [51], and Khelfaoui et al. (2022) [88] who postulate that ICT is an infant mortality-reducing factor.

The results also show that a percentage increase in income per capita reduces the U5MR by 0.1–0.4% at a 5% significance level. This entails that a growing GDP per capita can help reduce poverty levels and favor better nutrition and adequate housing, which ultimately improves health capital. Therefore, families have more resources to invest in their children’s health, education, living conditions, and well-being [47]. The result corroborates with Makuta and O’Hare (2015) [23], Boachie and Ramu (2016) [89], Akinlo and Sulola (2019) [5], Dhrifi (2020) [41], and Logarajan et al. (2022) [39], who postulate that GDP per capita is associated with low under-five mortality rates. The mediating impact of institutional quality on the U5MR-health expenditure nexus is presented in the following section.

Table 3 shows that the institutional quality variable is positively signed. Thus, a 1% increase in institutional quality is associated with a 0.03–0.1% increase in the U5MR at a 10% significance level. This result is inconsistent with the expected a priori of the study. The positive effect could be attributable to weak institutional quality, which challenges the healthcare service delivery [40] and contributes to high child mortality. The following section shows the results of the impact of institutional quality on the U5MR-health expenditure nexus in SSA countries in Table 4.

The Role of Institutional Quality

Table 4 presents the empirical results of the impact of institutional quality on the under-five mortality rate and health expenditure relationship. The main focus of the results is on interaction terms in the last eight rows. Thus, Model 1 is the benchmark model that presents the institutional quality’s effect on the U5MR-public health expenditure as a percentage of GDP (PHE) nexus. Models 2 to 4 show results of the impact of institutional quality on the U5MR-health expenditure as measured by domestic private health expenditure (HEDP), out-of-pocket health expenditure (HEOUP), and external health expenditure (HEEXT), respectively.

Table 4.
Results of the Role of Institutional Quality on the U5MR-Health Expenditure Nexus.

	(Model_1)	(Model-2)	(Model_3)	(Model_4)
	(SGMM)	(SGMM)	(SGMM)	(SGMM)
Variables	lU5MR	lU5MR	lU5MR	lU5MR
L.U5MR	0.00878 ***	0.00404 *	0.00860 ***	0.00621 ***
	(0.00180)	(0.00234)	(0.00113)	(0.00195)
dlphiv	0.179	0.241	1.035 ***	−0.264
	(0.596)	(0.401)	(0.343)	(0.308)
dlpun	0.116	0.0432	0.427 ***	0.117
	(0.0827)	(0.0749)	(0.139)	(0.0835)
limm	0.0588	−0.0800	0.230	−0.159 *
	(0.153)	(0.145)	(0.232)	(0.0919)
lmam	0.425 ***	0.589 ***	0.404 ***	0.495 ***
	(0.0871)	(0.0879)	(0.0959)	(0.0866)
lsenr	0.207 *	−0.0283	−0.0654 *	−0.153
	(0.109)	(0.0878)	(0.0323)	(0.147)
dlict	−0.00629	−0.0225	−0.206 **	−0.0252
	(0.0272)	(0.0281)	(0.0863)	(0.0265)
dlgdppc	−0.134 *	−0.0915	−0.123	−0.118 **
	(0.0744)	(0.250)	(0.0808)	(0.0513)
linst	0.344 ***	−0.378 **	0.0237	0.170 **
	(0.0922)	(0.1636)	(0.0470)	(0.0756)
Interaction Terms
lphe	−0.00921
	(0.0632)
lphelinst	−0.127 ***
	(0.0432)
lhedp		−0.0442
		(0.0616)
lhedplinst		0.103 *
		(0.0551)
dlheoup			−0.0854 *
			(0.0490)
dlheouplinst			−0.0164
			(0.0274)
lheext				−0.0416 *
				(0.0217)
lheextlinst				−0.0306 *
				(0.0180)
Constant	1.273 *	1.488 ***	3.034 *	1.291 ***
	(0.5234)	(0.3980)	(1.537)	(0.3490)
Observations	327	327	327	327
Number of countries	46	46	46	46
Number of instruments	26	27	28	26
F-Statistics p-value	0.000	0.000	0.000	0.000
AR(1) p-value	0.183	0.218	0.040	0.469
AR(2) p-value	0.102	0.515	0.100	0.141
Sargan test p-value	0.100	0.864	0.110	0.978
Hansen test p-value	0.873	0.706	0.567	0.439

Model 1 in Table 4 indicates the coefficient of the interaction term of public health expenditure (PHE) with institutional quality and external health expenditure (HEEXT) with institutional quality are negatively signified. The negative coefficient of the interaction terms indicates that institutions increase the negative effect of public and external health expenditure on the U5MR. This result is consistent with Makuta and O’Hare (2015) [23], Dhrifi (2020) [41], Langnel and Buracom (2020) [26], Dianda and Ouedraogo (2021) [47], and Hadipour, Delavari and Bayati (2023) [40], who indicate that improving institutional quality contributes to the development of democratic and meritocratic systems and efficient taxes which improves administrative capacity, thereby enhancing access to public and external health care and reducing infant mortality. This implies that public and external health expenditures are more likely to be effectively utilized in well-functioning healthcare systems with solid institutions, improving child health outcomes in SSA countries. Model 2 in Table 4 shows that the direct effect of institutional quality becomes negative after considering the interaction variable of domestic private health expenditure and institutional quality. Thus, a 1% increase in institutional quality reduces the U5MR by 0.4% at a 5% significance level. This implies that strong institutional quality can encourage greater healthcare utilization. This result aligns with Ouedraogo, Dianda and Adeyele (2020) [45], who suggest that institutional quality is relevant in improving health outcomes in the SSA region. Model 3 shows a marginal negative effect of out-of-pocket health expenditure on the U5MR.This implies that an increase in out-of-pocket health expenditure is associated with a decrease in U5MR, ceteris paribus. This aligns with Kimani (2014) [90] who suggests that out-of-pocket health expenditure helps households to restore health.

The interaction term of HEDP with institutional quality is positively signified. This implies that institutional quality enhances the positive impact of domestic private expenditures on the U5MR rate. The positively signed interaction term coefficient could be due to poor and weak institutional quality [35,41], which may lead to chronic underinvestment in public healthcare systems and a lack of healthcare infrastructure. This deficiency forces individuals and families to turn to private healthcare providers, incurring substantial out-of-pocket expenses for what they perceive as better-quality care [90,91]. This can leave marginalized communities with no choice but to rely on private providers, even if it leads to a financial burden, worsening health outcomes, and higher mortality rates.

[ad_2]