Child Mortality: what might be the underlying factors?

Introduction

Child mortality, defined here as the under-5 mortality rate per 1,000 live births, is considered a key indicator of a society's overall development and well-being. Over the past decades, global child mortality rates have declined significantly. However, substantial differences still exist  across countries. Developed countries usually  have lower mortality rates, while many low-income countries continue to face high child mortality due to weaker healthcare infrastructure and limited public resources. These disparities are unlikely to be random.

While working on a research project, I asked myself: What factors lead developed countries to have lower child mortality rates compared to low-income countries? After reading several articles, I identified a set of probable causes based on my own reasoning. I then used a statistical method called OLS regression to test how strongly each of these factors were correlated with child mortality rates. The following analysis presents the factors I found to be most strongly associated with child mortality.

GDP per capita

The article by O'Hare et al. (2013) shows the relationship between national income, GDP per capita and child mortality. Drawing on pooled international data and employing elasticity estimates, their study finds a significant inverse relationship between GDP per capita and infant mortality, with a pooled elasticity of −0.95. This means when GDP per capita increases by 10% the  infant mortality rate decreases by 5%. They pointed out that in order to reduce the child mortality rate, the government should not focus only on government health expenditure, but also address poverty.

Healthcare Expenditure and Medical Resources

Owusu et al. (2021) study directly examines the impact of government health expenditure  on child mortality.  They analyse cross-national data with a focus on income group differences, finding that increases in healthcare expenditure have a significantly greater mortality-reducing effect in developing countries than in high-income ones. The resource  suggests that both public and private investment in healthcare accessibility and affordability yields measurable improvements in health outcomes. Moreover,  Ayipe and Tanko (2023) focus specifically on low-income countries (in Sub-Saharan Africa), using panel data and a Fixed Effects regression model. Their results show a significantly negative relationship between government health expenditure  and under-five mortality (Fixed Effects = −5.275; p < 0.05), suggesting that by each 1% increase in government health expenditure, the under-5 mortality rate is reduced by 5.3%.

Corruption as an Institutional Factor

Beyond GDP per capita  and Government health expenditure, corruption is another significant factor that influences the under 5 mortality rate. The study from Hanf et al. (2011) analysed the political dimension of child mortality by examining the role of corruption. The study shows a multiple regression model with a strong explanatory fit (adjusted R² = 0.89). Their cross-national study finds that the Corruption Perception Index (CPI) is a statistically significant predictor of under-five mortality as the p-value is smaller than 0.01. Each 1 decrease in the corruption index creates  0.0644 decreases in the log of national under-five mortality. This finding underscores that even where government health expenditure and GDP per capita are sufficient, corruption can undermine their effective allocation.

Data method

Besides the articles I also  collected variables from the WDI and Corruption Perceptions Index data from 2000 to 2022 including all the factors above and one more: physician per 1,000 people (docs)   I did this in order find how heavily these factors influence the mortality rate per 1,000 live births (mort5). Variables are:  1. GDP per capita in constant 2015 US dollars (gdp_pc), 2. physicians per 1,000 people (docs), 3. domestic general government health expenditure per capita in PPP-adjusted current international dollars (gghed_pc), and 4. corruption index.

OLS regression and its analysis

The OLS results indicate that GDP per capita, government healthcare expenditure, physician density, and corruption are all negatively associated with under-five mortality. Government healthcare expenditure exhibits one of the largest coefficient magnitudes, suggesting a particularly strong relationship with mortality reduction. The model explains approximately 86% of the variation in child mortality (R² = 0.86), indicating strong explanatory power.

Conclusions and recommendations

Based on these findings, when aiming to reduce the child mortality rate, a government should be developomg its economic policy and increasing its GDP per capita.  For low-income countries where both GDP per capita and Government Healthcare Expenditure per capita are lower , the priority should be increasing gghed_pc (Government Healthcare Expenditure per capita) and ensuring it reaches frontline services. According to the result based on the OLS model, that gghed_pc (Government Healthcare Expenditure per capita) has a stronger and more consistent effect than (GDP per capita) alone, suggesting that targeted public investment in healthcare can reduce child mortality even in the absence of broad economic growth. At the same time, based on the data from OLS, expanding the physician workforce in rural areas could be another effective way to reduce the child mortality rate. A government should br preventing and reducing its corruption rate and allocating more resources to increase the physician rate per 1000 people. 

To conclude, research and data analysis suggests that economic and political factors, such as GDP per capita, national income levels, Government Healthcare Expenditure per capita, number of physicians per 1000 people and the level of corruption, are factors which play a significant role not only in the survival of infants but also in  affecting health outcomes and mortality rates in general.