Economic development and healthcare expenditure are two correlated trends that determine child survival in Africa. The increase in economic output will enhance income, nutrition, and educational and medical services, which subsequently will reduce child mortality (19, 20). Nevertheless, the extent of this advantage is determined by the fairness of income gains distribution and the availability of growth dividends to vulnerable groups, which is supported by evidence on the significance of inclusive development (21). Health spending, in its turn, has a greater impact on child survival by investing in infrastructure, preventive health, and service delivery infrastructure (22, 23). However, the impact of such spending can be minimized in case of mismanagement of the public resources or institutional inefficiency (24). All of these results suggest that as economies grow, mortality can be indirectly decreased, but in the long term, it is health investment that can be highly effective and efficient only when it is well-governed. Overall, sustainable reductions in child mortality require an integrated approach in which economic growth is complemented by well-governed and efficiently financed health systems. Hence, we would expect that both economic progress and health expenditure play crucial roles in improving child survival outcomes across African countries.

One of the most chronic dangers to child survival is environmental degradation especially in third world countries where rate of urbanization and loose regulatory measures are the leading cause of high rates of pollution. Carbon emissions and air pollution in particular are detrimental and put pressure on respiration, weak birth weight, and premature death (10, 11). Use of biomass fuel, poor consumption of energy and congestion in urban areas are other factors that expose children to toxic pollutants in African contexts (8). Poverty is also strengthened by environmental stress, which restricts access to nutrition and healthcare, as well as increases health susceptibility in infants and young children (9). It has been indicated that the trend to improve the air quality and climate change can offset the anticipated health benefits with regards to economic growth even (15). Therefore, environmental degradation—particularly air pollution—must be understood not merely as an ecological concern but as a critical public-health challenge that undermines sustainable development and endangers child survival. Hence, we hypothesize that higher levels of air pollution are associated with increased child mortality across African countries.

The spatial and demographic transitions are significant to the outcome of child survival in Africa. Child mortality continues to be preeminently caused by high fertility, especially because of resource dilution, decreased maternal healthcare access, and decreased parental investment per child (12, 13). To facilitate a reinforcing feedback loop, there is a tendency to increase mortality to drive up fertility in order to create a demographic trap that makes the population health increase more slowly (25, 26). On the other hand, urbanization may facilitate the survival of children by enhancing access to healthcare, educational infrastructures, as well as sanitation amenities (27, 28). That being said, unplanned urban growth can increase the risk of being exposed to pollution, communicable disease, and overcrowded living conditions, thus undoing possible health benefits (29). The experience in African settings thus indicates that though high fertility has a constant propensity of increasing the risk of mortality, urbanization when properly controlled can reduce demographic tensions. The interaction of these aspects is one of the factors that lead to the spatial differences present in child survival throughout the continent. Accordingly, we expect fertility to exacerbate, and urbanization to alleviate, child mortality across African countries.

Building on the health production framework and the existing empirical literature, child mortality is viewed as an outcome shaped by economic capacity, public health investment, environmental quality, and demographic structure. Economic growth enhances household income and government fiscal space, thereby improving access to nutrition and healthcare services, while public health expenditure strengthens preventive and curative health systems. Conversely, environmental degradation and demographic pressure impose significant health burdens by increasing disease exposure and constraining household-level health investments. In the African context, these effects are expected to be heterogeneous, reflecting differences in baseline vulnerability, institutional capacity, and access to essential services across countries. Accordingly, the following hypotheses are proposed:

Figure 1 demonstrates hypothesis development of the study.

Although numerous studies have examined the determinants of child mortality, most have treated economic, environmental, and demographic factors as isolated influences rather than as interacting systems. Previous research has typically emphasized income growth, fertility, or health expenditure individually, overlooking how these dimensions collectively shape child survival outcomes across the African continent. Critical gaps therefore persist in understanding the simultaneous and conditional effects of these variables, particularly in resource-constrained settings where multiple risks coexist (2, 4, 7). First, limited research has explored how economic growth and health expenditure interact under conditions of high fertility and environmental deterioration. While evidence suggests that stronger economies and increased health investments reduce mortality (6, 19), there remains uncertainty regarding whether these benefits persist when pollution and fertility pressures rise concurrently. Existing cross-country analyses often rely on mean-based models that neglect heterogeneity in mortality intensity, thereby obscuring disparities between low- and high-mortality nations (1, 15).

Second, environmental degradation, particularly air pollution, has been recognized as a critical health risk, yet few studies have integrated it with socioeconomic and demographic dynamics in the African context. Research indicates that high levels of particulate matter and carbon emissions increase child deaths (8, 10, 11), but most environmental studies remain detached from development modeling. This fragmentation prevents a thorough understanding of how pollution may offset the health gains achieved through economic growth or greater healthcare investment. Furthermore, empirical work only rarely considers the moderating role of urbanization, which can either alleviate or intensify mortality risk depending on governance and infrastructure quality (12–14). Third, comparative regional analyses often limit their focus to sub-Saharan Africa, creating a gap in continent-wide evidence that covers the collective African Union framework. As a result, there is limited understanding of how regional policy heterogeneity, economic structure, and demographic transition jointly influence child survival across Africa’s northern, western, eastern, central, and southern subregions (3, 4).

To address these limitations, the present research provides the first continent-wide empirical assessment of the joint and interactive effects of economic growth, health expenditure, urbanization, fertility, and air pollution on child mortality in 53 African Union countries from 2000 to 2024. By adopting a multidimensional perspective, this study reveals the asymmetric and conditional nature of mortality determinants that previous analyses have overlooked (2, 15). The approach demonstrates how the benefits of economic expansion and urban growth can be constrained by high fertility and environmental stress in low-income contexts. Empirically, the study contributes by employing a distribution-sensitive analytical framework that captures variation across the mortality distribution, identifying structural vulnerabilities among high-mortality countries that traditional mean-based estimations conceal (1, 4). Theoretically, it situates the child mortality debate within an integrated sustainability paradigm, linking economic prosperity, environmental health, and demographic transition, thereby expanding the scope of development and public-health literature in Africa (9, 11). Overall, this research fills a critical gap by offering a comprehensive, evidence-based understanding of how multiple development drivers interact to shape child survival outcomes. It advances the discourse on achieving Sustainable Development Goal 3.2 within Africa’s unique demographic and environmental context and underscores the necessity for coordinated African Union policies that align economic growth with environmental protection and health system resilience.

The present study uses the concept of balanced panel data comprising 53 member countries of the African Union from 2000--2024. The data combines data on the World Development Indicators (WDI), the World Health Organization (WHO), and our World in Data so that there is consistency, coverage, and comparability of the data over time and across nations. This period was selected according to the time of the heightened health and environmental reforms in Africa, such as the Millennium Development Goals (MDGs) and the shift to the Sustainable Development Goals (SDGs). Applying this analysis to the year 2024 enables the consideration of post-pandemic recovery processes that can potentially transform the economic–health–environment nexus (2, 15).

Child mortality (CM) will be used as the dependent variable, measured as the number of deaths of children under five per 1,000 live births. It serves as a useful proxy for assessing the extent of child health and overall population well-being within a country, which has been widely used in the existing literature (30, 31). In addition, economic growth (GDP per capita), air pollution (CO emissions per capita), health spending (as a percentage of GDP), urbanization (urban population share), and the fertility rate (births per woman) are the explanatory variables. These variables are based on empirical evidence that has suggested relationships between income, pollution, demographic pressures, and mortality (7, 9, 11). The empirical setting is Africa since it has the highest rate of under-five mortality globally and is highly heterogeneous in terms of institutional, environmental, and demographic characteristics (4, 32). Table 1 summarizes the variable definitions and the source and measurement units.

This theoretical framework is based on the conceptual provisions of the health output (HO) model created by Grossman (33) and views health as an outcome of many other inputs, such as income, education, quality of the environment, and access to healthcare. The Grossman model initially concentrated on micro-level individual behavior, in which health is a type of capital that individuals can invest in through preventive and curative efforts. On this basis, later authors such as Majeed and Ozturk (34) and Siddique and Kiani (35) extended the HO model to macroeconomic factors in which national and environmental conditions coexist and shape the health status of societies. This paper further extends the HO framework to the African context by defining child mortality as a negative measure of health output, while population-related (fertility, urbanization) and environmental (air pollution) characteristics are treated as inputs into the health production process. The modified framework assumes that health outcomes improve with rising economic growth and with increased investment in healthcare but deteriorate under demographic pressure and environmental degradation.

Mathematically, the relationship can be expressed in Equation 1:

The extended health production relationship now can be write in Equation 2:

This study investigates the relationship between child mortality and a set of economic, environmental, and demographic factors across African Union member states. The model extends the macro-level health output framework to include the principal variables discussed in Section 2, namely, economic growth, health expenditure, air pollution, fertility, and urbanization. Let i = 1, …, 53 index African Union member countries and t = 2000, …, 2024 index years. The dependent variable CM_it denotes under-five child mortality (deaths per 1,000 live births). The regressors include income GDP_it (GDP per capita), air pollution CO₂_it (carbon dioxide emissions per capita), health expenditure HE_it (percent of GDP), urbanization URP_it (urban population share), and fertility FR_it (births per woman). Then Equation 3 formulated as follows:

Equation 4 applies a natural logarithmic transformation to all variables. Equation 4 can be written as follows:

Where, β 1 to β 5 denote coefficients of explanatory variables, ε it used as error term and β 0 shows intercept. Taking logarithms allows the coefficients to be interpreted as elasticities, indicating the percentage change in child mortality resulting from a one-percent change in each explanatory variable. The log-linear specification also helps stabilize variance, reduce heteroskedasticity, and better capture nonlinear relationships among variables across diverse African economies.

Table 2 clarifies the theoretical pathway linking macroeconomic, environmental and demographic factors to child mortality, and directly supports the empirical specification adopted in the study.

A series of diagnostic tests was applied before the actual implementation of the main estimations to ensure that the econometric model would be valid and robust. The variance inflation factor (VIF) was initially used to test the possibility of multicollinearity existing between the explanatory variables and thus ensure that the correlations were within the critical range and therefore did not affect the estimates of the coefficients. Subsequently, cross-sectional dependence (CSD) was tested via the Pesaran (36) test, which revealed intercountry connections due to shared regional shocks such as health crises and economic spillovers. Pesaran and Yamagata (37) proposed a slope heterogeneity test that was then utilized to test structural differences among countries. The unit root tests Levin–Lin–Chu (LLC), cross-sectional Im-Pesaran–Shin (CIPS), and augmented Dickey–Fuller (ADF) were used to determine the order of integration (38–40). These gave a combination of I(0) and I(1) series. The relationships of variables in the long run were also established with the help of Westerlund and Edgerton (41), Pedroni (42), and Kao (43) cointegration tests.

To approximate the heterogeneous determinants of child mortality, the method of moments quantile regression (MMQR) by Machado and Silva (17) is used in the present study. The MMQR records distributional heterogeneity in that it permits the variation of coefficients across quantiles, with more detailed insight into how economic, environmental, and demographic variables influence mortality at lower and higher levels of burden. Complementary estimators were employed to prove the strength of the MMQR estimates. The Driscoll–Kraay standard error (DKSE) methodology eliminates cross-sectional dependence, heteroskedasticity, and serial correlation, which guarantees unbiased standard errors. Moreover, the endogeneity, which occurs as a result of reverse causality between the economic variables and health variables, was controlled with the aid of system GMM and difference GMM estimators. These models are able to guarantee consistent and efficient estimation by employing internal instruments (lagged regressors) and specification tests such as AR (1), AR (2), Sargan and Hansen. The consistency of the findings of the MMQR, DKSE and GMM approaches corroborates the reliability and stability of the findings in explaining child mortality in African countries (Figure 2).

The findings of this study provide substantial theoretical contributions to the disciplines of health economics, sustainable development, and demographic studies. Anchored in the health output framework of Grossman (33), the results reaffirm that health outcomes are not the result of a single economic or social determinant but rather emerge from the interaction of multiple economic, environmental, and demographic forces. In this sense, health is best viewed as a multidimensional stock of human capital produced jointly by household behavior, public investment, and environmental conditions (1, 4).

By extending the Grossman model to the macro level, this research conceptualizes child mortality as a negative outcome of health production—reflecting the structural capacity of nations to sustain human development. Unlike traditional growth–health frameworks that predominantly rely on income-driven explanations of health, the present study integrates economic growth, environmental quality, fertility dynamics, health expenditure, and urbanization into a single analytical system. This multidimensional framework aligns with contemporary sustainability theories that emphasize the interdependence of economic prosperity, environmental stewardship, and institutional performance (9, 15).

The evidence that fertility and pollution accelerate child mortality even under conditions of economic expansion challenges traditional growth-centric perspectives (6, 11). Such findings suggest that income gains alone are insufficient for sustainable health improvements if environmental degradation and high fertility persist. The results thus support the emerging consensus that economic progress can only translate into better population health when accompanied by demographic transition and ecological preservation (2, 7). Theoretically, this implies a composite model of health production in which environmental stressors and demographic pressures act as counterforces to the benefits of growth and health investment.

Furthermore, the quantile-based estimation reveals critical distributional heterogeneity in the determinants of child mortality. The study finds that the effects of economic growth, healthcare expenditure, fertility, and air pollution vary systematically across country-specific mortality levels. This variation indicates that the health production frontier is non-linear and conditional on structural and institutional thresholds, a notion consistent with Zakaria et al. (4) and Gayawan et al. (15). In practical terms, the marginal returns to income or health spending are greater in high-mortality countries with weaker institutions and higher demographic burdens. This insight extends Grossman’s original formulation by introducing a threshold-sensitive dimension to the macro-economic determinants of health. Hence, this study reaffirms that achieving sustained improvements in child health requires an integrated framework that simultaneously advances economic capacity, strengthens institutional and demographic systems, and mitigates environmental degradation, thus redefining health output as a multidimensional outcome of sustainable development.

The results have important managerial and policy implications for governments, regional bodies, and development partners. Given the heterogeneity revealed through the findings of the present study, interventions should be differentiated across country groups. For high-mortality, high-fertility countries, where fertility pressures and environmental stressors strongly dampen the benefits of income expansion, policy priority should be directed toward rapid fertility decline via universal reproductive health services, expanded family planning, and girls’ secondary education, consistent with the demographic transition pathway outlined by Amegbor and Addae (2) and Rahman and Alam (7). Strengthening primary healthcare, maternal-child services, and immunization networks is particularly relevant in fragile settings where health spending shows the highest marginal return, as also evidenced by Rahman et al. (1). In these contexts, performance-linked health financing and conditional public transfers may accelerate the mortality response.

For high-pollution emerging economies, where income and urbanization contribute to mortality reduction yet are offset by environmental degradation, coordinated energy–environment governance is essential. Investments should prioritize clean household energy, industrial emission control, and air-quality monitoring, aligning with sustainability arguments in Gayawan et al. (15) and Hassan and Murad (9). Ministries of health, energy, and environment must jointly design pollution-responsive child health interventions that expand access to clean energy technologies while scaling neonatal and respiratory care. Meanwhile, lower-mortality transitioning countries, where demographic burdens are declining and health systems expanding, require policies that consolidate gains. Here, the focus should shift toward surveillance-based mortality tracking, urban planning, and digital health systems that maintain downward mortality trends, echoing institutional and threshold effects highlighted by Zakaria et al. (4). At the regional level, the African Union, AfDB, and UNECA should adopt monitoring frameworks under the Africa Health Strategy (2023–2030) to classify countries into the three categories and tailor financing windows accordingly. South–South cooperation with emerging economies in Asia and Latin America can further support technology transfer in telehealth, renewable energy, and demographic governance. Local implementation should involve civil-society actors, youth groups, and community health workers to strengthen accountability, enhance reproductive autonomy, and reinforce public awareness around environmental risks. Through such interventions, progress toward SDG 3.2 becomes more feasible and equitable across Africa’s heterogeneous mortality landscape.

Although the present study provides robust empirical and theoretical insights into the determinants of child mortality in Africa, several limitations should be acknowledged. First, while the use of MMQR and GMM estimations effectively addresses issues of endogeneity, heterogeneity, and distributional asymmetry, the approach remains fundamentally linear and may not fully capture complex threshold, feedback, or spatial spillover effects inherent in the development–health–environment nexus. Future studies should therefore incorporate dynamic spatial econometric frameworks or nonlinear models, such as panel threshold regressions, Bayesian hierarchical models, or machine-learning algorithms, to account for multidirectional causality among key variables like fertility, pollution, and income.

Second, the study period (2000–2024) reflects the contemporary era of health and environmental reforms in Africa but does not include earlier decades of post-independence development, structural adjustment, and pre-MDG health transitions. Historical data covering the 1970s through the 1990s could enable researchers to trace the long-term evolution of health systems and socioeconomic disparities that laid the foundations for present-day child mortality trends. Incorporating such datasets would allow for longitudinal analyses that capture the continuity and structural transformation of African health and environmental policies over time. Third, the study’s cross-national analytical design, although suitable for capturing general patterns across 53 African Union countries, may overlook unique contextual conditions at the sub-regional, national, and local levels. The contextual specificity of mortality determinants, shaped by political stability, cultural norms, and resource endowments, requires more nuanced country- or community-specific analyses. Future research could therefore focus on case studies or mixed-methods approaches that combine econometric modeling with field surveys, interviews, or administrative health data to better understand the interplay of policies and behavioral dynamics at the micro level.

Finally, comparative evidence across regions beyond Africa—such as Asia, Latin America, or the Middle East—would enhance the external validity and generalizability of the results. Cross-regional studies could help assess whether the identified distributional heterogeneities in mortality determinants are unique to Africa’s demographic and environmental conditions or represent a broader developing-world phenomenon. Overall, future research should aim to integrate richer, multidimensional datasets with innovative analytical tools to reveal the complex socio-environmental, institutional, and policy mechanisms that shape child health. Such methodological and contextual extensions would deepen the understanding of Africa’s mortality dynamics and provide more tailored policy insights supporting the Sustainable Development Goals, particularly SDG 3.2, in diverse development contexts.