📈 economics

Beyond Income: The Independent Contributions of Health Expenditure and Digital Infrastructure to Population Longevity

This paper analyzes data from 261 countries to demonstrate that government health expenditure and digital infrastructure independently and significantly drive population longevity and reduced child mortality, challenging the notion that income alone is the primary determinant of life expectancy.

Original authors: H. Semih Yildirim

Published 2026-07-10

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: H. Semih Yildirim

Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine the world as a giant, global race track. For decades, runners born in wealthy nations like Japan have been finishing the race of life at age 84, while runners born in places like Sierra Leone have been finishing around age 62. That's a gap of more than two decades, compressed into the sheer luck of where you were born.

For a long time, the big question was: "What makes the difference?" Most people assumed the answer was simple: Money. The idea was that if a country just got richer, its people would automatically live longer. It was like thinking the only way to win a race is to buy the fastest car.

But this new study, looking at data from 261 countries over 25 years (from 2000 to 2024), throws a wrench into that idea. It suggests that while money matters, it's not the only driver, and it's not even the most important one on its own.

The Big Reveal: It's a Three-Legged Stool

The researchers used a special math trick (called LMG variance decomposition) to see how much each factor actually contributes to the "longevity gap." Think of it like a pie chart of life expectancy.

Here is the surprising slice of the pie:

Government Health Spending: This accounts for 23.0% of the difference in how long people live across countries.
National Income (GDP per capita): This accounts for 22.4%.
Digital Infrastructure: This accounts for 21.3%.

The paper explicitly argues against the idea that income is the "king" of longevity. In this joint model, government health spending and national income are statistically tied for first place. They are "co-leaders." Digital infrastructure is right there with them, a "near-peer."

The study also rules out the idea that Out-of-Pocket (OOP) spending (money families pay directly for care) is a major independent driver of longevity. Even though high OOP costs are terrible for poor families and create poverty traps, in this specific math model, it only explains 3.6% of the difference between countries. Why? Because countries where families pay less out-of-pocket are usually already richer and better governed, so the "OOP" number is just a shadow of those other bigger factors.

The "Digital" Secret Sauce

One of the most playful and surprising parts of the study is the role of digital infrastructure. You might think "digital" means fancy apps or telemedicine. But the study suggests the real magic is in the pipes and wires.

The researchers found that the Telecommunication Infrastructure Index (TTI)—basically the strength of cell towers, internet cables, and network connections—is the biggest part of digital infrastructure that helps kids survive.

The Analogy: Imagine a health system as a delivery service.

Government Health Spending is the money to buy the vaccines and medicine.
Digital Infrastructure (TTI) is the delivery truck and the GPS.

If you have a truck full of vaccines (money) but no roads or GPS (digital infrastructure), the vaccines sit in a warehouse and rot. The study suggests that digital infrastructure acts as an "efficiency multiplier." It doesn't replace the need for health spending; it makes that spending work better. It ensures vaccines reach the cold chains, supplies get to remote clinics, and community health workers can stay connected.

The "Time Travel" Problem

Here is where the study gets very specific about what it can't prove yet.

The study found that when a country increases its government health spending by 10%, the very next year, infant mortality drops by about 0.8%, and under-5 mortality drops by 0.75%. This is a clear, measurable win for kids.

However, when they looked at adult life expectancy, the results were "statistically inconclusive." The paper doesn't say health spending doesn't help adults; it says the effect is too slow to see in this specific dataset.

The Analogy: Think of health spending like planting an oak tree.

Child mortality is like the sapling. You water it (spend money), and you see it grow taller the next year.
Adult life expectancy is the full-grown oak. It takes decades of consistent watering to see the tree reach its full height.

Because the study only looked at a 25-year window, it couldn't see the full growth of the oak tree. The "lag" (the time it takes for the investment to pay off) for adults is likely longer than the study could measure. So, the paper suggests that while we see immediate results for children, the results for adults are likely there, just hidden in the long-term future.

What This Means for the Future

The study concludes that we can't just wait for a country to get rich before we expect people to live longer. We also can't just throw money at health systems without fixing the digital "pipes" that deliver the care.

The authors suggest that Income, Health Spending, and Digital Infrastructure are like three legs of a stool. If you remove one, the whole thing wobbles. They seem to work best together, not as a sequence where you get rich first and fix health later.

The Bottom Line:
The paper doesn't claim to have "solved" the mystery of longevity. It admits that proving cause-and-effect is hard and that more research is needed to see the long-term effects on adults. But it does provide a strong, measured benchmark: To close the 20-year gap in life expectancy, countries need to treat health budgets and digital networks as equally important partners, right alongside economic growth. It's not just about having the money; it's about having the system to spend it effectively and the wires to deliver it.

Technical Summary: Beyond Income: The Independent Contributions of Health Expenditure and Digital Infrastructure to Population Longevity

Problem and Motivation
The paper addresses a fundamental empirical question in global health economics: which government investments independently contribute to population longevity, and how do these contributions compare when income, governance, and digital infrastructure are considered simultaneously? While the "Preston curve" tradition establishes a robust link between national income and life expectancy, and the Filmer-Pritchett debate highlights the primacy of institutional quality over spending levels, existing literature rarely quantifies the independent contributions of health investment, income, governance, and digital infrastructure within a unified framework.

A critical gap exists in the literature regarding digital infrastructure. While digital systems theoretically improve health logistics and administrative capacity, macro-level evidence on their independent contribution to longevity—distinct from their correlation with income and governance—is sparse. Furthermore, standard bivariate analyses are misleading due to severe multicollinearity among predictors (government health expenditure, GDP per capita, governance quality, and digital readiness all correlate strongly, $r = 0.78–0.92$ ). The paper seeks to resolve whether health expenditure is merely a proxy for underlying development conditions or if it possesses independent leverage in explaining cross-national longevity differences.

Methodology
The study utilizes a comprehensive global panel dataset covering 261 countries from 2000 to 2024 (6,525 country-year observations), integrating data from the World Bank WDI, WHO GHO, UN EGDI, and Worldwide Governance Indicators. The empirical strategy employs two complementary approaches to address different identification challenges:

LMG Variance Decomposition (Primary Analysis):
- Objective: To attribute cross-country variation in life expectancy to each predictor simultaneously, regardless of entry order.
- Method: The Lindemann–Merenda–Gold (LMG) method calculates the average marginal $R^2$ increment for each of the six predictors (log government health expenditure, OOP share, log GDP per capita, government effectiveness, EGDI, and internet users) across all $p!$ (720) possible orderings.
- Significance: This ordering-invariant approach overcomes the bias of sequential hierarchical partitioning, providing a precise measure of the unique explanatory power of each variable when all others are jointly controlled.
Two-Way Fixed Effects (TWFE) Panel Regressions (Secondary Analysis):
- Objective: To estimate within-country associations between changes in health investment/digital infrastructure and longevity outcomes over time, absorbing time-invariant country heterogeneity and global shocks.
- Specification: $Y_{it} = \alpha_i + \gamma_t + \beta X_{i,t-1} + \delta'Z_{it} + \varepsilon_{it}$ , where $X$ represents lagged government health expenditure or EGDI.
- Controls: The model includes contemporaneous time-varying controls (GDP per capita, government effectiveness, rule of law, urbanization, physician density) and fixed effects for country and year.
- Lag Structure: Regressors are lagged one year to partially address reverse causality (e.g., spending increases in response to mortality crises) and to align with the temporal structure of health production (procurement to service delivery).

Key Contributions

Systematic Joint Decomposition: This is the first study to apply an ordering-invariant variance decomposition (LMG) to jointly quantify the independent contributions of health investment, income, governance, and digital infrastructure to cross-country longevity variation.
Re-evaluating the Hierarchy of Determinants: The paper challenges the implicit hierarchy that treats digital investment as secondary to direct health spending and questions the dominance of income in explaining longevity gaps once digital infrastructure is accounted for.
Differentiating Short-Run vs. Long-Run Dynamics: By separating child mortality outcomes from adult life expectancy, the paper documents that within-country health spending responsiveness is concentrated in child survival (shorter lag structures) rather than adult longevity (which requires multi-decadal cumulative investment), providing a grounded basis for calibrating policy expectations.

Results

Cross-Sectional Variance Decomposition (LMG):
- In a six-predictor model explaining 74.6% of life expectancy variance, government health expenditure (23.0%) and GDP per capita (22.4%) are statistically indistinguishable co-leaders.
- Digital Infrastructure (EGDI) accounts for 21.3% of the variance, statistically indistinguishable from GDP but carrying a modestly smaller share than health expenditure.
- Government effectiveness (15.1%) and internet users (14.6%) are secondary predictors.
- Out-of-Pocket (OOP) spending contributes only 3.6%, suggesting its theoretical importance as a financial protection variable is largely captured by its correlation with income, governance, and public spending levels.
- Bootstrap inference confirms that health expenditure and EGDI each retain unique explanatory power not captured by the other predictors.
Within-Country TWFE Estimates:
- Child Mortality: A 10% increase in lagged per capita government health expenditure is associated with approximately 0.8% lower infant mortality, 0.75% lower under-5 mortality, and 0.9% lower neonatal mortality in the following year. These associations remain robust across specification ladders.
- Adult Longevity: Associations with adult life expectancy and Healthy Life Expectancy (HALE) are statistically inconclusive (negligible coefficients), attributed to the long cumulative lag structures of adult survival that exceed the identification window of a 25-year panel with a one-year lag.
- Digital Infrastructure Mechanism: Sub-index analysis reveals that the Telecommunication Infrastructure Index (TTI) is the primary driver of child mortality reductions, outperforming Online Service (OSI) and Human Capital (HCI) indices. This supports a "logistics channel" hypothesis where connectivity improves vaccine supply chains and community health worker reach.
- OOP Share: No statistically significant within-country association was found between OOP share and longevity outcomes, likely due to the slow-moving nature of financing structures within countries over a 25-year horizon.
Income Group Heterogeneity:
- Income-group splits for life expectancy are underpowered and inconclusive. However, the negative point estimates for Lower-Middle-Income Countries (LMC) align with the pooled child mortality results, suggesting health spending has visible near-term returns in settings where public health systems are the primary delivery mechanism for vaccination and maternal care.

Significance and Claims
The paper claims to provide a precise descriptive benchmark for the global longevity gradient. It argues that the race to close the two-decade longevity gap is not solely a race for economic growth but is equally a race to build effective health systems and productive digital infrastructure.

Methodological Significance: The study demonstrates that partial models omitting digital infrastructure risk misattributing variance. The comparable independent explanatory power of health spending, income, and digital infrastructure suggests they should be treated as jointly important inputs rather than a sequential hierarchy.
Policy Implications:
- Budget Prioritization: Health expenditure and digital infrastructure are presented as complements rather than substitutes. Digital infrastructure (specifically physical connectivity) acts as an efficiency multiplier for health spending, potentially raising the return on health budgets.
- Sequencing: For health system efficiency, prioritizing physical connectivity (mobile towers, broadband) may be more consequential for child survival than administrative digitization platforms in the short term.
- Expectation Management: Policymakers should expect observable population-level returns from health investments primarily in child mortality outcomes within a one-to-three-year horizon, while adult longevity improvements reflect longer-term cumulative investments.

The paper concludes that while causal identification requires quasi-experimental designs (e.g., fiscal shocks, natural experiments), the current descriptive evidence establishes that health investment, national income, and digital infrastructure together account for approximately two-thirds of the explained cross-country variance in longevity, with no single input dominating the others.

The Big Reveal: It's a Three-Legged Stool

The "Digital" Secret Sauce

The "Time Travel" Problem

What This Means for the Future

Technical Summary: Beyond Income: The Independent Contributions of Health Expenditure and Digital Infrastructure to Population Longevity

More like this