BACKGROUND: Government health care spending (GHS) is of increasing importance to child health. Our study determined the relationship between reductions in GHS and child mortality rates in high- and low-income countries.
METHODS: The authors used comparative country-level data for 176 countries covering the years 1981 to 2010, obtained from the World Bank and the Institute for Health Metrics and Evaluation. Multivariate regression analysis was used to determine the association between changes in GHS and child mortality, controlling for differences in infrastructure and demographics.
RESULTS: Data were available for 176 countries, equating to a population of ∼5.8 billion as of 2010. A 1% decrease in GHS was associated with a significant increase in 4 child mortality measures: neonatal (regression coefficient [R] 0.0899, P = .0001, 95% confidence interval [CI] 0.0440–0.1358), postneonatal (R = 0.1354, P = .0001, 95% CI 0.0678–0.2030), 1- to 5-year (R = 0.3501, P < .0001, 95% CI 0.2318–0.4685), and under 5-year (R = 0.5207, P < .0001, 95% CI 0.3168–0.7247) mortality rates. The effect was evident up to 5 years after the reduction in GHS (P < .0001). Compared with high-income countries, low-income countries experienced greater deteriorations of ∼1.31 times neonatal mortality, 2.81 times postneonatal mortality, 8.08 times 1- to 5-year child mortality, and 2.85 times under 5-year mortality.
CONCLUSIONS: Reductions in GHS are associated with significant increases in child mortality, with the largest increases occurring in low-income countries.
What’s Known on This Subject:
After the recent economic recession, policy interventions including austerity measures led to reductions in government spending on health care in many countries. However, there is limited research into the effects of changes in government health care spending on child health.
What This Study Adds:
Reductions in government health care spending are associated with long-lasting adverse effects on child health globally, especially in low-income countries. Given pressures to diminish health expenditures, we caution that reduced spending should be achieved through increased efficiency of care delivery.
The World Health Organization’s fourth Millennium Development Goal (MDG 4) aimed to reduce under-5 mortality by two-thirds between 1990 and 2015.1 However, although progress has been made, recent analyses estimated that only 10 countries of 67 with high child mortality (≥40 deaths per 1000 live births) were on track for meeting the MDG 4 target.2 Studies have tried to address key components pertinent to aiding progress on maternal and child health. These include health care interventions (for example, skilled birth attendants and measles vaccination) and socioeconomic inequity measures in middle- to low-income countries,3 as well as increased coverage of high-impact child survival interventions including insecticide-treated bed nets, nutritional interventions, care seeking, and treatment of childhood diarrhea and pneumonia.4
Economic fluctuations contribute to periods of prosperity and austerity together with changes in key areas of government funding, namely: health, education, and infrastructure.5,6 The effects of economic crises on population health have been explored in single groups or small groups of high-income countries,7–13 with a broad range of health outcomes including child health being adversely affected.14–16 After the recent global recession and policy interventions including austerity measures, government health care spending (GHS) has decreased in several countries.17 The role of GHS has also been the focus of cost-control debates in developed countries.18
To date, however, little research has explored the effect of changes in GHS on population health globally. In light of the recent movements to extend the MDG initiative, and calls for a renewed child survival strategy and longer-term targets for 2035,19 it is pertinent to determine how economic fluctuations, and more specifically, changes in GHS may affect child mortality.
We sought to determine the effect of reductions in GHS on child mortality rates worldwide between 1981 and 2010, comparing these relationships in high- and low-income countries.
Economic data were extracted from the World Bank’s World Development Indicators, 2013 edition.17 Child mortality data were obtained from the Institute for Health Metrics and Evaluation.19 Data were collectively available for 176 countries. The definitions of child health outcomes, GHS (measured as a percentage of total government spending), and high- and low-income countries are described in Supplemental Table 4.
We used multivariate regression analysis to measure the associations between child mortality (dependent variable) and changes in GHS (independent variable). All decreases in government spending on health care were incorporated into the analysis, and the change in mortality associated with a 1% decrease in GHS was calculated. To ensure that the results were not confounded by extreme variations in certain countries, we used a fixed-effects approach in the regression models, including 176 dummy variables for the 176 countries included in the analysis. This approach meant that models evaluated mortality changes within individual countries while holding constant time-invariant differences between countries, including higher predispositions to child mortality and political, cultural, and structural differences. We also controlled for country population demographics by introducing controls for the total population size, percentage of the population older than 65 years, and percentage of the population younger than 15 years. Using the Cook–Weisberg test,20 the data were found to be characterized by heteroskedasticity (where subsamples have different distributions). Robust standard errors were therefore included in the regression models, accounting for the heterogeneity in the dataset, for example, due to differences in the way that countries measured government spending. This methodology has been widely used in health-economic studies and is regarded as a statistically robust approach.7,21
Our basic linear fixed-effects statistical model uses the following equation: where H is the health metric (child mortality); U is the measure of government health spending; α represents the population structure of the country being analyzed; D is a dummy variable for each country included in the regression model; and ε is the error term.
We conducted 1- to 5-year time-lag multivariate analyses to evaluate the longer-term effects of a 1% decrease in GHS in a single year on child mortality. We also categorized countries as high and low income using the definitions in Supplemental Table 4, and we repeated the multivariate regression analyses using the same model for each of these categories. Robustness checks were performed to ensure validity of results (see Results).
We used Stata SE version 12 for the analysis (Stata Corp, College Station, Texas).
Child Mortality Data
Data were collectively available for 176 countries (Supplemental Table 5), equating to a population of ∼5.8 billion as of 2010. Table 1 shows the results of multivariate regression analyses for the 176 countries worldwide between 1981 and 2010. A 1% decrease in GHS was associated with a significant increase in child mortality rates; neonatal (regression coefficient [R] 0.0899, P = .0001, 95% confidence interval [CI] 0.0440–0.1358), postneonatal (R = 0.1354, P = .0001, 95% CI 0.0678–0.2030), 1- to 5-year (R = 0.3501, P < .0001, 95% CI 0.2318–0.4685), and under 5-year (R = 0.5207, P < .0001, 95% CI 0.3168–0.7247) mortality.
We conducted time-lag analyses to assess the effect on child mortality 1 to 5 years after a 1% decrease in GHS in a single year (Table 1, Fig 1). All 4 child mortality measures consistently displayed significant deteriorations up to 5 years after the decrease in GHS. The magnitude of the regression coefficient decreased as the time from the change in GHS elapsed.
Countries were categorized as high and low income in accordance with World Bank definitions (Supplemental Table 4). Compared with high-income countries, low-income countries displayed larger, significant deteriorations in all 4 child mortality measures (Table 2, Fig 2). Specifically, after a 1% decrease in GHS, low-income countries experienced greater deteriorations of ∼1.31 times neonatal mortality; 2.81 times postneonatal mortality; 8.08 times 1- to 5-year mortality, and 2.85 times under-5 mortality, compared with high-income countries (Table 2).
A series of further analyses were performed to evaluate the robustness of our results (Table 3). To isolate the impact of the changes in GHS, we introduced additional economic controls, adjusting for inflation, gross domestic product (GDP) per capita, government debt as percentage of GDP, and real interest rates; decreases in GHS continued to be associated with an increase in all mortality measures. Second, we introduced infrastructure controls, adjusting for urbanization, access to water, and calorie intake; decreases in GHS continued to be associated with significant increases in all mortality measures except postneonatal mortality. Third, we introduced infectious disease controls to account for child HIV prevalence, tuberculosis treatment success rates, and diphtheria/pertussis/tetanus immunization rates; decreases in GHS continued to be associated with significant increases in all mortality measures. After introducing all economic, infrastructure, and infectious disease controls simultaneously into our model, decreases in GHS continued to be associated with significant increases in all mortality measures. Fourth, we controlled for out-of-pocket health expenditures; decreases in GHS continued to be associated with significant deteriorations in all metrics except neonatal mortality. Fifth, we controlled for private health care spending; decreases in GHS continued to be associated with a significant increase in all mortality measures. Finally, we repeated the analysis using an alternative data source for child health outcomes: mortality measures from the World Bank (Table 3), which uses 3 mortality categories: neonatal, infant, and under-5.17 Decreases in GHS were also associated with significant increases in all 3 mortality measures.
Our analysis has demonstrated an association between GHS and child mortality at the global level. Specifically, we have shown that a 1% decrease GHS is associated with significant increases in child mortality across 176 countries worldwide. The association is significant for up to 5 years after decreases in GHS and remains after controlling for economic, infrastructure, infectious disease, out-of-pocket expenditure, and private health spending indicators. Low-income countries experienced the largest changes in child mortality, in some cases up to 8 times those of the high-income countries.
The mechanisms by which reductions in GHS increase child mortality are not widely explored in the current literature. In most countries, the public sector plays a significant role in providing health and education services.5,6 Countries use different measures to limit GHS at times of austerity and in the context of cost control. Some attempt to enforce pay cuts to health care workers; others constrain the amounts spent on pharmacological interventions and medical equipment.6,18 Other strategies used to reduce the health care payment burden are to increase out-of-pocket expenses, by increasing patient or employer premiums and introducing user charges.18 Non–price rationing of health care (for example, increases in waiting times) is also used to reduce health care costs.18 All of these measures potentially diminish access to health care and the quality of care delivered.
Reduced GHS also results in lower hospital budgets, which in turn reduces the availability of hospital resources,22 including reductions in the quantity of health workers.18 For example, from 1975, the long-lasting economic crisis led to a decline in health expenditures, imports of medical supplies and drugs, and salaries for physicians (resulting in them leaving the country); under-5 mortality increased for almost 20 years.24 Prematurity is a leading cause of newborn deaths and under-5 mortality; up to two-thirds of these deaths can be prevented if skilled health professionals perform effective health measures at birth and in the first week of life.25 Public health campaigns to improve the recognition of illness and the promotion of effective management of infections, birth asphyxia, and preterm or low-birth-weight babies require health care funding.26 Higher-technology interventions such as assisted ventilation for neonates are also extremely costly.26 For the postneonatal period, the leading causes of death include pneumonia and diarrhea.1 Reduced GHS may reduce the availability of relevant medications, such as antibiotics, and supportive oxygen therapy.4,25
We found the increased child mortality to be significant up to 5 years after a reduction in GHS. Although in our study we have not explored the effects beyond 5 years, it is not inconceivable that reductions in health care spending may have long-term effects on the economy of a nation. Specifically, reduced health care spending may reduce GDP per capita, as (1) less healthy populations are less productive economically, and (2) reduced investments in health care and weaker efforts toward preventive measures can result in worsened outcomes, development of more severe disease, and requirements for more expensive care.27 Therefore, reductions in GHS may culminate in greater costs to a nation’s economy.28
Our analysis has also shown that cuts in GHS adversely affect child health to a greater scale in low-income countries. More than 70% of all under-5 deaths occur in Africa and southeast Asia, with a background of poverty, suboptimal care seeking, and weak health care systems.29,30 This difference is likely because the lowest-income countries may have preexisting weak health systems, which suffer most when GHS is reduced, as key components of basic health care may be lost, costing substantially more lives than in high-income countries, where effective health systems may be more resilient to budgetary cuts.31–33 Low-income countries are also more vulnerable to economic shocks, as they are more likely to have national debt and strong dependence on a single production good.24 Resource scarcity has led to the need for innovations in health systems to sustain effective responses and improvements in health; however, weak systems in low-income countries may hinder the receptiveness to innovation.34 The literature focusing on the health effects of economic fluctuations and government expenditures in low-income countries is limited. One study in Uganda demonstrated that increased health care expenditures, especially on vaccination, have a positive impact on infant mortality. Increasing the rate of vaccination to 100% was considered the most influential and probably most cost-effective intervention, reducing infant deaths by 16 per 1000 births.35 Other high-impact child survival interventions shown to quickly reduce mortality include appropriate treatment of childhood diarrhea and pneumonia.4 Our study demonstrates the need for additional research into the mechanisms by which economic factors affect child mortality in low-income countries, to encourage policies and interventions that will effectively reduce preventable child deaths.
Finally, caution must be taken in debates concerning cost control and budget restrictions of health care. Increases in health care spending do not always translate to clinically better outcomes; the United States has one of the highest health care expenditures per capita but does not appear to be efficient and effective in meeting the health needs of children.36 In comparison, the United Kingdom has achieved significantly better clinical outcomes with lower health care expenditures.36 It is important to note that if budgetary cuts are implemented without simultaneous efficiency gains, or sufficient support in child health care, rises in child mortality may occur. A recent article in The Lancet emphasized that reductions in inequality are important in reducing child mortality in high-income countries.37 Although we propose that substantial additional investments will be required to improve child health beyond the MDG target of year 2015, policymakers should note that a health care system must translate increases in spending to significant improvements in population health.
Our study has several limitations. First, our analysis was performed on 176 countries at a global level; thus our results reflect a collective effect rather than country- or regional-specific trends. Nevertheless, we have conducted subanalyses comparing the strength of associations between decreases in GHS and increases in child mortality in high- and low-income countries. Second, the retrospective observational nature of our study does not allow us to draw inferences about causality. Third, effects of changes in economic factors including changes in government expenditure may be delayed and long lasting;33 we analyzed only the short-term effects up to 5 years after the reduction in GHS, and further longitudinal analysis may be useful. Fourth, we have not modeled intrayear variations in mortality; however, by looking at whole years, we eliminate spurious trends that might be seen as seasonal in nature. Fifth, although we focus on the quantity of government spending on health care, we do not factor how efficiently or effectively these funds were spent. Indeed, it is entirely feasible that a country may spend less on health care and achieve greater outcomes due to efficient application of funds or avoidance of ineffective or inappropriate interventions. Finally, we have not been able to classify the type of GHS, as these data are not publicly available from the World Bank.
Nevertheless, our study used worldwide data, taken from high-quality, objective, and centralized databases, avoiding selection and recall bias. The volume of data used allowed for high statistical power and multiple robustness checks. By focusing on countries at a global level over a 30-year period, this study permitted consideration of macroscopic trends. Notably, our study used a fixed-effects multivariate regression analysis model. This model, along with the multiple robustness checks, accounts for many criticisms leveled at other studies looking at the relationship between health outcomes and economic changes: namely, the omission of potential confounders that are likely to be correlated with both GHS and child mortality. By using this panel-data approach, we also controlled for time-invariant heterogeneity between countries, something that aggregate, time-series analyses fail to do. In addition, all data used are publicly accessible, supporting the reproducibility of our study.
Our results have shown that reductions in GHS are associated with long-lasting increases in child mortality at a global scale, especially in low-income countries. Given plans to extend the MDG initiative, it is important to appreciate that economic changes, in particular changes in government spending, may affect child health outcomes. Given the risk of reductions in health expenditures, we caution that reduced spending should be achieved through increased efficiency of care delivery; otherwise, child outcomes may worsen.
We thank Dr. Emily Prior.
- Accepted December 30, 2014.
- Address correspondence to Ka Ying Bonnie Ng, Obstetrics and Gynaecology Department, Chelsea and Westminster Hospital, London, United Kingdom. Email:
Dr Maruthappu and Mr Williams extracted the data; Drs Maruthappu and Ng drafted the initial manuscript; and Dr Maruthappu, Dr Ng, Mr Williams, Prof Zeltner, and Prof Atun conceptualized and designed the study, interpreted the data, reviewed and revised the manuscript, and approved the final manuscript as submitted.
FINANCIAL DISCLOSURE: The authors have indicated they have no financial relationships relevant to this article to disclose.
FUNDING: No external funding.
POTENTIAL CONFLICT OF INTEREST: The authors have indicated they have no potential conflicts of interest to disclose.
- ↵World Health Organization. MDG 4: reduce child mortality. Geneva, Switzerland: WHO; 2013. Available at: http://www.who.int/topics/millennium_development_goals/child_mortality/en/. Accessed May 5, 2014
- ↵United Nations. The millenium development goals report. New York: UN; 2010. Available at: http://www.un.org/millenniumgoals/pdf/MDG%20Report%202010%20En%20r15%20-low%20res%2020100615%20-.pdf. Accessed January 2015
- Barr B,
- Taylor-Robinson D,
- Scott-Samuel A,
- McKee M,
- Stuckler D
- Ruhm C
- Tapia Granados JA
- ↵World Bank. World Development Indicators. Washington, DC: World Bank; 2013. Available at: http://data.worldbank.org/data-catalog/world-development-indicators. Accessed May 5, 2014
- ↵Mladovsky P, Srivastava D, Cylus J, et al. Policy summary 5. Health policy responses to the financial crisis in Europe. New York: WHO; 2012. Available at: http://www.euro.who.int/en/data-and-evidence/evidence-informed-policy-making/publications/2012/health-policy-responses-to-the-financial-crisis-in-europe. Accessed May 6, 2014
- Stuckler D,
- Basu S,
- McKee M,
- King L
- Martin A,
- Lassman D,
- Whittle L,
- Catlin A,
- National Health Expenditure Accounts Team
- Africa Progress Panel. Maternal health: investing in the lifeline of healthy societies and economies. Africa Progress Panel; 2010. Available at: http://www.africaprogresspanel.org/policy-papers/maternal-health-investing-in-the-lifeline-of-healthy-societies-economies/. Accessed May 16, 2014
- ↵World Health Organization. Child mortality and causes of death. Geneva, Switzerland: WHO; 2014. Available at: http://www.who.int/gho/child_health/mortality/en/. Accessed May 5, 2014
- ↵Amiri A, Gerdtham UG. Impact of maternal and child health on economic growth: new evidence based on Granger causality and DEA analysis. Geneva, Switzerland: Partnership for Maternal, Newborn and Child Health; 2013. Available at: http://www.ministerialleadershipinhealth.org/wp-content/uploads/sites/19/2013/03/Econ-benefits_econometric-study_Lund-University1.pdf. Accessed May 6, 2014
- Costello A,
- Osrin D,
- Manandhar D
- Stuckler D,
- Meissner C,
- Fishback P,
- Basu S,
- McKee M
- Atun R
- ↵Ssewanyana S, Younger S. Infant mortality in Uganda: determinants, trends and the Millennium Development Goals. J Afr Econ. 2008;17(1):34–61
- Infant and Child Mortality Estimates by Country 1970-2010. Seattle, Washington: Institute for Health Metrics and Evaluation; 2014. Available at http://ghdx.healthdata.org/record/infant-and-child-mortality-estimates-country-1970-2010. Accessed May 5, 2014
- World Bank. What Is the World Bank Atlas method? Washington, DC: World Bank. Available at: https://datahelpdesk.worldbank.org/knowledgebase/articles/378832-what-is-the-world-bank-atlas-method. Accessed May 5, 2014
- Copyright © 2015 by the American Academy of Pediatrics