blackpaper · the problem
700 million people live on less than $2.15 a day. That threshold is not a comfortable poverty, it is the line below which the World Bank considers survival itself uncertain: food, water, shelter, and basic health, in the cheapest countries on earth, still out of reach for one in twelve humans alive today (World Bank, Poverty and Inequality Platform, 2024).
The historical trend is real and it is good: the extreme-poverty share of humanity fell from 38% in 1990 to under 9% before 2020, the fastest sustained reduction in material deprivation in recorded history (Our World in Data / World Bank). But the curve bent the wrong way after 2020. COVID-19, inflation, and conflict pushed the count back up from roughly 650 million in 2019 to 700 million in 2022, and it has not fully recovered.
The remaining poverty is not evenly spread and it is not the same problem it was in 1990. It is concentrated: over 60% of people in extreme poverty now live in Sub-Saharan Africa, and the World Bank's own 2024 report concludes that at current rates, ending extreme poverty globally "could take more than a century" for the hardest-to-reach half of that population (World Bank, "Ending Poverty for Half the World Could Take More Than a Century," October 2024). Growth alone, the engine that did most of the work from 1990 to 2015, is decelerating in the low-income countries where poverty is now concentrated, and it is increasingly concentrated in fragile and conflict-affected states where growth, by definition, struggles to take hold.
What it costs the world right now: the World Bank estimates the total shortfall, the sum of every dollar needed to lift every person in extreme poverty exactly up to the $2.15/day line, at roughly $70 billion a year (World Bank, Poverty, Prosperity, and Planet Report 2024). That number understates the true cost of solving the problem, because a one-time cash top-up does not build the institutions, markets, and infrastructure that keep people out of poverty once they cross the line. But it is a hard floor: $70 billion a year is the theoretical minimum cost of closing the gap by pure income transfer, in a world whose gross output is over $105 trillion.
The severity for the people inside this line is close to total: extreme poverty crowds out nearly every other axis of wellbeing, health, education, physical safety, dominating daily decision-making for as long as it lasts. This is not a problem of unknown solutions. Cash transfers and targeted public-health interventions have been tested more rigorously than almost any other class of intervention in economics, repeatedly, at scale, with randomized controlled trials (Banerjee and Duflo, Nobel Prize 2019). The open question is not "does this work." It is why, with tools this well-proven and a bill this small relative to global GDP, 700 million people are still on the wrong side of the line.
whitepaper · the proposal
The tools to end extreme poverty are not missing. Direct cash transfers, distributed with minimal overhead through mobile money and biometric identity rails, are among the most rigorously evidenced interventions in the social sciences: hundreds of randomized controlled trials show consistent, durable gains in income, assets, and child outcomes (J-PAL; Banerjee and Duflo). GiveDirectly alone has moved over $1 billion in unconditional cash to people in extreme poverty since 2009, with transaction costs that fall as digital rails mature (GiveDirectly, public financials). The unlock now is not new science. It is delivery infrastructure: mobile money penetration in Sub-Saharan Africa crossed 33% of GDP in transaction value by 2023 (GSMA, State of the Industry Report on Mobile Money 2024), and biometric national ID systems, most visibly India's Aadhaar, have shown that hundreds of millions of unbanked people can be reached with verifiable, fraud-resistant digital payments at near-zero marginal cost per transfer.
What would actually have to be built: interoperable, low-fee payment rails between mobile money operators, aid agencies, and governments in the highest-poverty geographies (largely Sub-Saharan Africa and South Asia's remaining pockets); cheap, portable digital identity that works without a smartphone or reliable internet; and a shift in aid architecture away from in-kind and program overhead toward direct transfer, which the evidence increasingly favors on both cost-effectiveness and dignity grounds. None of this requires a technological breakthrough. It requires capital and distribution discipline applied to tools that already exist and are already proven.
The dollar figure, worked from first principles: the World Bank's Poverty, Prosperity, and Planet Report (2024) puts the total annual income shortfall for the ~700 million people below the $2.15/day international poverty line at approximately $70 billion a year, derived as roughly $100 per person per year in transfers (World Bank, 2024). That is the floor cost of closing the gap by transfer alone, and it is the number this market is actually priced against: it is 0.07% of global GDP (~$105 trillion, World Bank 2024), smaller than what OECD countries already pledge, on paper, in development assistance (the long-standing 0.7%-of-GNI commitment, which most donor countries do not meet).
That $70 billion a year is a floor, not a ceiling, because transfers alone do not build the growth, health systems, and institutions that keep a household above the line permanently. A fuller market includes the delivery rails that make transfers cheap and reliable at this scale: fintech and identity infrastructure for the roughly 1.4 billion adults still unbanked worldwide (World Bank Global Findex 2021, most recent full survey), for whom a functioning digital-payment relationship is itself worth building even independent of a specific transfer program. If even 10% of that unbanked population becomes a paying (or subsidized-fee) customer of a low-cost digital financial rail at $5-10/year in transaction margin, the delivery-infrastructure layer alone is a $700 million-$1.4 billion/year business, sitting underneath, and enabling, the larger $70 billion/year transfer market.
Confidence note, stated plainly: the $70 billion figure is a hard, well-sourced number from the World Bank's own poverty-gap methodology, high confidence on the arithmetic. What is genuinely uncertain, and this is a real open question rather than a hedge, is whether closing the gap by transfer converts into durable poverty exit, or whether the underlying growth and institutional deficits reassert themselves once transfers stop. The 2024 World Bank report's own headline, that ending poverty for the hardest half of this population could take over a century at current growth rates, is the honest counter-evidence: money alone has not been sufficient historically, even when it has been available. The buildable version of this thesis is not "write a $70 billion check." It is: build the distribution rails that make transfers, wage income from growth, and public services all cheaper to deliver to the same 700 million people, so that whichever lever moves fastest gets to actually reach them.