If You Could Know Only One Number About a Country, It Should Be Under-5 Mortality

Pick any country on Earth. You are allowed one statistic -- just one -- to understand its level of development. Not GDP per capita. Not literacy. Not life expectancy. The single most revealing number is the under-5 mortality rate: how many children out of every 1,000 born alive die before their fifth birthday.

That number is a compression of everything else. High under-5 mortality means inadequate healthcare, unsafe water, poor sanitation, insufficient nutrition, low maternal education, and weak institutions. Low under-5 mortality means a country has figured out the basics. It has clean water, vaccines, trained birth attendants, oral rehydration salts, and -- crucially -- educated mothers who know when to seek care. Under-5 mortality is not just a health statistic. It is a civilization diagnostic.

And by that measure, the last quarter-century has been one of humanity's finest hours. Global under-5 mortality fell from roughly 93 per 1,000 live births in 1990 to about 37 per 1,000 by 2020 -- a 60 percent decline. That translates to roughly 7 million additional children surviving to their fifth birthday every single year compared to 1990 rates. This is arguably the most important thing that has happened in the world during most people's lifetimes, and almost nobody talks about it.

But the progress is profoundly uneven. Nearly 5 million children still die before age 5 every year, and the geography of those deaths is concentrated with brutal precision.

The Curve

Plot every country's GDP per capita against its under-5 mortality rate and the pattern is immediate: richer countries have lower child mortality. But the relationship is logarithmic, not linear, and that shape matters.

Each bubble is a country. Size reflects population. Color reflects region. The dashed line is the fitted curve across 236 countries. The correlation between log-transformed GDP per capita and under-5 mortality is -0.79.

On the left side of the chart, the curve is steep. Going from $1,000 to $5,000 per capita (PPP) corresponds to a drop in under-5 mortality from roughly 90 to 40 per 1,000. That is 50 fewer children dying per thousand born -- purchased, in effect, by the basic infrastructure that early economic development brings: paved roads to clinics, electricity for vaccine refrigerators, clean water systems, and trained midwives.

On the right side, the curve flattens. Going from $30,000 to $75,000 per capita barely moves the needle. The United States, at $74,000 GDP per capita, has an under-5 mortality rate of 6.5 per 1,000 -- higher than Japan (2.4), Germany (3.7), or the United Kingdom (4.5). Money alone does not guarantee that children live.

The Outliers

The most interesting countries are those that deviate from the curve -- in both directions.

Sri Lanka is the standout overperformer. With a GDP per capita of just $13,000 (PPP), it achieves an under-5 mortality rate of 6.1 per 1,000 -- comparable to European nations spending three to five times as much. Sri Lanka invested in free universal healthcare and universal education starting in the 1930s, long before most developing countries. The payoff has compounded for decades. Its 21.8-point residual below the fitted curve is one of the largest in the world.

Rwanda sits 13.7 points below the curve. In 2000, a country still recovering from genocide had under-5 mortality of 185 per 1,000 -- nearly one in five children dying before their fifth birthday. By 2023, that number had fallen to 40. Rwanda achieved this through a community health worker program that placed trained volunteers in every village, combined with aggressive immunization campaigns and performance-based financing for health facilities. The decline of 78 percent in just over two decades is among the fastest ever recorded.

Ethiopia shows a similar trajectory. Under-5 mortality fell from 140 per 1,000 in 2000 to 47 by 2023 -- a 67 percent decline. Ethiopia's Health Extension Programme, which deployed 38,000 community health workers to rural areas starting in 2003, is widely credited for this progress.

On the other side, Nigeria is the most dramatic underperformer. Nigeria's full economic profile shows how oil wealth has failed to translate into basic health outcomes. At $7,800 GDP per capita, the fitted curve predicts under-5 mortality around 37 per 1,000. The actual rate is 105. That residual of +68 represents tens of thousands of preventable deaths every year. Nigeria has the world's highest absolute number of under-5 deaths -- roughly 860,000 per year. The country has oil wealth but has failed to translate it into basic health infrastructure, particularly in the northern states where immunization rates remain below 30 percent and maternal education is limited.

Equatorial Guinea is even more extreme in relative terms. With oil-driven GDP per capita of $15,700, its predicted under-5 mortality should be around 25. The actual rate is 71. The country demonstrates what happens when resource wealth accrues to a narrow elite and never reaches the health system.

The Progress

The slope chart below shows 30 countries' under-5 mortality in 2000 versus the most recent available year. Each line connects the two data points. Green lines indicate countries that achieved more than 50 percent decline. Yellow indicates 30-50 percent. Red indicates less than 30 percent.

The steepest declines are concentrated among countries that started worst. Sierra Leone went from 224 to 94. Niger from 228 to 115. Mali from 188 to 91. These are still unconscionably high numbers -- but the direction is unmistakable. Countries that were losing one in five or one in four children are now losing closer to one in ten.

China's progress is remarkable: from 37 to 6.2, an 83 percent decline that moved a quarter of the world's children into a zone of survival comparable to Southern Europe. India achieved a 70 percent decline, from 92 to 28. Bangladesh, starting at 85, reached 31 -- a 64 percent drop driven by one of the most effective community health systems in the developing world.

The laggards are telling. Somalia managed only a 40 percent decline. Nigeria, despite its wealth, achieved just 42 percent. Pakistan, held back by gender inequality, conflict, and weak governance in frontier regions, managed 46 percent. And Cuba, starting from an already low base of 8.6, saw essentially no change -- because it was already at the floor achievable with its resources.

What Drives Survival

If income alone explained child mortality, there would be nothing for policy to do except grow GDP. But the data tells a more nuanced story. We computed the correlation between under-5 mortality and five potential drivers across all countries with available data.

Female adult literacy is the single strongest correlator at -0.87 -- stronger than income itself (-0.79). Educate women, and children survive. This is not a subtle statistical relationship. It is the most robust finding in the entire field of development economics. Literate mothers recognize danger signs in sick children, seek care earlier, follow treatment instructions more accurately, practice better hygiene, and space their pregnancies more widely. Every additional year of female schooling in a developing country reduces under-5 mortality by an estimated 5-10 percent.

DPT immunization rate (-0.57) and measles immunization (-0.63) are both moderately strong. These capture two things simultaneously: the direct protective effect of vaccines, and the presence of a functioning health system capable of delivering them. A country that can get three doses of DPT vaccine to 90 percent of its one-year-olds has supply chains, cold chains, trained health workers, and community trust. These are the same capabilities needed to deliver oral rehydration therapy, treat pneumonia, and manage complicated births.

Education spending as a percentage of GDP shows the weakest correlation (-0.20). This is not because education does not matter -- the female literacy result proves it does. The weak correlation reflects the fact that spending percentages are a poor proxy for actual educational outcomes. Some countries spend heavily but inefficiently. Others, like Vietnam, achieve remarkable educational outcomes at modest cost.

The Unfinished Work

The global average conceals a geographic concentration that should be uncomfortable. Sub-Saharan Africa accounts for roughly 29 percent of global births but more than 50 percent of under-5 deaths. The region's average under-5 mortality remains above 70 per 1,000 -- approximately ten times higher than Europe or East Asia.

Within Sub-Saharan Africa, the crisis is further concentrated. Five countries -- Nigeria, the Democratic Republic of the Congo, Ethiopia, Tanzania, and Niger -- account for more than half of all under-5 deaths on the continent. Nigeria alone accounts for nearly one-quarter.

South Asia, the other historically high-mortality region, has made dramatic progress. India's decline from 92 to 28 per 1,000 between 2000 and 2023 is one of the most consequential public health achievements in human history, given its population scale. Bangladesh has done even better in relative terms. But Pakistan remains a stubborn outlier at 59 per 1,000, held back by conflict zones in the northwest, deeply rooted gender inequality that limits maternal education, and a polio eradication program that has been undermined by political instability and distrust.

The causes of under-5 death are known and, in most cases, treatable. Pneumonia, diarrhea, malaria, neonatal complications, and malnutrition account for the vast majority. Oral rehydration salts cost pennies. Insecticide-treated bed nets cost a few dollars. The pneumococcal vaccine exists. The interventions are not waiting to be invented. They are waiting to be delivered.

Success Stories: What They Did Right

Bangladesh is perhaps the most instructive case. A densely populated country with limited natural resources and frequent natural disasters, Bangladesh has achieved health outcomes far beyond what its income would predict. The BRAC organization -- the world's largest NGO -- built a parallel health delivery system that reaches virtually every village. Community health workers trained in basic triage and oral rehydration therapy became the backbone of child survival. Female education expanded rapidly, with girls' secondary enrollment rising from 14 percent in 1990 to over 70 percent by 2020. Contraceptive use increased, birth rates fell, and children born into smaller families received more resources per capita.

Rwanda took a top-down approach. Explore Rwanda's economic data to see how a low-income country turned governance into survival. After the genocide, the government rebuilt the health system from scratch with a focus on community-based delivery. Every village received a pair of trained community health workers -- one male, one female. A performance-based financing system rewarded health facilities for measurable outcomes rather than inputs. Health insurance coverage expanded from 7 percent in 2003 to over 90 percent by 2015 through a community-based insurance scheme (Mutuelle de Sante). The result: the fastest decline in under-5 mortality in Africa.

Ethiopia deployed 38,000 Health Extension Workers -- young women with one year of training -- to rural communities starting in 2003. These workers provided immunizations, family planning, malaria treatment, and health education at the village level. Ethiopia's under-5 mortality decline of 67 percent since 2000 demonstrates that a low-income country can make extraordinary progress with a coherent primary care strategy and political commitment.

The Story the Numbers Tell

Under-5 mortality is declining almost everywhere. The global rate fell by about 55 percent between 1990 and the early 2020s. That trajectory, if sustained, would bring the world close to the Sustainable Development Goal target of 25 per 1,000 by 2030 -- though most of Sub-Saharan Africa will miss it.

The factors that drive this progress are not mysterious. Female education. Immunization. Clean water. Oral rehydration therapy. Trained birth attendants. Community health workers. Family planning. These interventions are proven, affordable, and scalable. The countries that deployed them -- regardless of income level -- saved children. The countries that did not -- regardless of income level -- lost them.

This is the most successful global development story of the past half-century. It is also unfinished. Five million children dying before their fifth birthday every year is not an acceptable number. But thirty years ago, the number was twelve million. The trajectory is extraordinary. The remaining challenge is concentrated, known, and solvable. It requires political will, health system investment, and -- above all -- the education of girls.

If you could change only one thing to save children's lives, the data is clear: educate their mothers.

---

Methodology

Raw data inputs (all from the World Bank World Development Indicators, joined at the country-year level inside DuckDB):

• Mortality rate, under-5 (per 1,000 live births) — originally compiled by the UN Inter-agency Group for Child Mortality Estimation (UNICEF, WHO, World Bank, UN DESA)
• GDP per capita, PPP (constant 2021 international $)
• Government expenditure on education, total (% of GDP)
• Immunization, DPT (% of children ages 12-23 months)
• Immunization, measles (% of children ages 12-23 months)
• Literacy rate, adult female (% of females ages 15 and above)
• Population, total

Derived formulas:

# Log-linear income-mortality fit (ordinary least squares)
u5_mortality_hat = a * ln(gdp_pc) + b
residual         = u5_mortality - u5_mortality_hat
# overperformer if residual < -10, underperformer if residual > +10

# Progress between 2000 and latest available year (>= 2015)
pct_decline = 100 * (u5_2000 - u5_latest) / u5_2000

# Factor correlation (Pearson) across all countries with non-null values
r(factor, u5) = corr(factor_i, u5_mortality_i)

Time period: 1990-2023. For the scatter plot and factor analysis, we use the most recent year per country where both under-5 mortality and GDP per capita are available. For the slope chart, we compare year-2000 values to the latest available year (2015 or later).

Sample: 236 countries after excluding aggregates and countries missing either under-5 mortality or GDP data. Cuba is notably absent from the scatter plot because the World Bank does not publish GDP PPP data for Cuba.

Curve fitting: The logarithmic fit (under-5 mortality = a x ln(GDP per capita) + b) yields a = -17.78, b = 196.35, with a correlation of -0.79. Residuals measure deviation from the curve: positive residuals indicate higher mortality than predicted by income (underperformers), negative residuals indicate lower mortality (overperformers).

Factor correlations: Pearson correlation coefficients between under-5 mortality and each factor, computed across all countries with non-missing values for that factor. Female literacy has the strongest absolute correlation (-0.87), followed by log GDP per capita (-0.79), measles immunization (-0.63), DPT immunization (-0.57), and education spending as % GDP (-0.20).

Limitations: GDP per capita PPP does not capture income distribution -- inequality within countries can mean that health spending does not reach the poorest. Female literacy data has significant gaps (many high-income countries do not report it because rates are near 100%). The cross-sectional correlations do not establish causation, though the causal mechanisms for female education and immunization are well-established in the literature. Under-5 mortality data for conflict-affected countries (Somalia, South Sudan, Syria) may be estimated rather than measured.