You already have the data. The framework encodes the wrong causal model.
You track development outcomes for 185 countries. Your reporting systems contain life expectancy, child mortality, children per woman, income, and educational completion — by country, by year, going back decades.
Inside that data is a pattern. Every country that crossed the development threshold — life expectancy above 69.8 and fewer than 3.65 children per woman, the United States' own values in 1960 — did so through education. Not after becoming wealthy. Not after building health systems. Through education, while still poor.
Bangladesh crossed both thresholds in 2014 at $1,159 per person. Cuba crossed in 1974. Korea started from one of the lowest incomes in the world in 1953 and crossed within one generation.
The pattern is not hidden. It is in the data you already collect. The problem is that the framework you built to interpret the data encodes the wrong causal model.
Your institution created the Human Development Index around three dimensions: health, education, and income. Then it adopted the Sustainable Development Goals and ranked one of those three dimensions fourth — behind hunger, which the HDI does not include. Goal 1 is poverty. Goal 2 is hunger. Goal 3 is health. Goal 4 is education. The framework contradicts itself. Nobody noticed for a decade.
That ordering is not neutral. It is a theory of development — the same theory that has structured international aid for four decades: address poverty and health first, then education follows.
OECD DAC data shows the result: health receives 2–3 times more international aid than education. Within education aid, secondary receives a fraction of primary. Within secondary, girls' completion in the lowest-baseline countries — where the return is highest — receives the least.
Every low-income finance minister understands this incentive structure. Health programmes attract donors. Vaccination campaigns produce reportable results within a funding cycle. Secondary school construction in rural areas does not. The minister optimises rationally against what the framework rewards — and the framework rewards the wrong thing.
You did not create this incentive to be perverse. But the ordering of goals is the theory of change, and every downstream actor — bilateral donors, national planners, implementing agencies — reads the ordering as priority. SDG 4 is goal number four. It is funded like goal number four.
The great health achievements of the MDG era — the fall in child mortality, the rise in life expectancy across the developing world — are real. They are also misattributed.
WHO, UNICEF, GAVI, PEPFAR — these programmes are staffed by educated workers deploying welfare technology to populations whose educational level has been rising since the 1950s. The global expansion of secondary education created the substrate on which health delivery operates. Educated mothers navigate health systems. Educated communities adopt public health practices. Educated workforces staff the clinics.
China's barefoot doctor programme is the canonical example. Dreze and Sen (1989) credit it for China's life expectancy gains. But when life expectancy is matched on mean years of schooling rather than income, China was below education-matched peers from 1965 to 1991. The barefoot doctors were not producing exceptional outcomes; China was underperforming what its education level predicted. When Deng dismantled the barefoot doctor system after 1980, the LE convergence rate was statistically unchanged (p = 0.78) — because the educational base built into rural households did not leave with the doctors.
The health gains are real. They do not compound on their own. They persist only as long as external delivery continues — unless the population is educated enough to sustain them independently. Education gains are structurally irreversible: a generation of educated parents cannot be uneducated. The framework funds the reversible gains and underfunds the irreversible mechanism.
This is not about bad intentions. It is about time horizons.
A vaccination campaign shows results in twelve months. A secondary education programme shows results in twenty-five years — one generation forward. No bureaucrat's career spans a generational lag. No funding cycle rewards what cannot be measured before the next replenishment round.
The reporting framework rewards what is measurable in a funding cycle. Rational career behaviour — delivering visible, attributable results within the reporting period — produces the wrong allocation. This is not corruption. It is optimisation against the wrong objective function.
But the problem is deeper than time horizons. Every professional in the system has a domain. The health specialist needs health interventions to be the mechanism. The economist needs growth to be the mechanism. The governance expert needs institutions to be the mechanism. If education is the fundamental cause, then every other domain is a downstream effect — real, valuable, but not the cause. No professional can accept that their life's work addresses a consequence rather than a cause. This is not a flaw in the people. It is a structural feature of any system staffed by specialists: each specialist has a career-long stake in their domain being primary.
The framework must change so that rational behaviour produces the right allocation. That is your job. Not to ask people to be less rational, but to change what rationality optimises for. And not to ask specialists to abandon their domains, but to reorder the framework so that the mechanism comes first and the downstream effects are understood as what they are: the things educated people produce.
A girl who completes lower secondary education:
This is not a projection. It is the measured sequence across 185 countries. The question is not whether secondary education for girls produces these outcomes. It does. The question is whether the international funding architecture is structured to deliver secondary education first — or whether it continues to fund the outcomes directly, leaving the mechanism unfunded.
The countries with the largest positive departures from their generational baselines — Maldives, Cape Verde, Bhutan, Tunisia, Nepal, Vietnam, Bangladesh (companion paper Table 4, DOI 10.5281/zenodo.19641662) — all had one thing in common: sustained political commitment to female secondary education above what their economic baseline would have produced automatically. None achieved this through health spending alone. None achieved this through income growth alone. All made education the priority when they were poor, and income and health followed.
The SDG framework will be revised. That revision is the leverage point.
Reorder the lead indicator. Secondary completion rates — not enrollment, completion — should be the primary tracking metric. A country's female lower-secondary completion today predicts its life expectancy and children per woman in 25 years more reliably than its current health spending does. Tie disbursements to education. Health aid conditionality should require concurrent education investment — not sequential. A country receiving health assistance should be expanding secondary access simultaneously, because the health gains will not sustain themselves without the educational base. Shift within education. Primary enrollment is nearly universal. The binding constraint is now secondary completion, especially for girls. Within education aid, the weight must move from primary enrollment to secondary completion in countries below 40% female completion — where the demographic headwinds are strongest and the return is highest. Track the generational rate. Report educational expansion in percentage points per year alongside health metrics. Korea expanded at 2.13 pp/yr. Bangladesh at 1.30. India at 0.87. That rate determines whether a country crosses the development threshold in one generation or three.The data is already in your system. The pattern has been visible for decades. The countries that crossed did so through education. The countries that are still waiting are the ones where the framework directed attention and capital elsewhere.
You do not control national budgets. You do not control private capital. You control something more powerful: the framework that tells every national planner and every donor what to measure, what to fund, and what success looks like.
The 2030 Agenda is mid-term. There is time to change the operating theory without changing the goals. The evidence is available, the historical cases are documented, and the mechanism is well-identified across 185 countries.
You put education fourth. They got the message.