The Surprisingly Small Price Tag to End Extreme Poverty: $318 Billion a Year

It’s easy to assume the world’s biggest problems are too large or distant for ordinary people to fix. But a new study using AI-enhanced, highly detailed data finds that completely eliminating extreme poverty could cost about $318 billion per year — roughly 0.3% of global GDP.

The report models targeted, no-strings-attached cash transfers to those in the most severe need so they can afford basic necessities: food, shelter and medicine. While $318 billion sounds large in absolute terms, the authors stress it is modest compared with routine global spending: for example, Americans will spend more than three times that on holiday shopping this year, and global annual spending on alcohol is roughly seven times larger.

How The Study Works

Rather than relying on an idealized "poverty-gap" calculation — which would theoretically cost about $30 billion a year by topping each household up exactly to a poverty line — the researchers use administrative and survey data that many low-income governments already collect. They apply AI-influenced analytics to estimate operationally feasible transfer amounts. That makes the proposal implementable at scale, even if some households would receive slightly more or less than the mathematically perfect amount.

"It’s not that it’s prohibitively expensive to address extreme poverty," says Joshua Blumenstock of UC Berkeley, a co-author. "It’s institutionally and politically difficult."

What The Transfers Would Achieve

Using a $2.15-per-day threshold (the World Bank adjusted its extreme-poverty line to $3 earlier this year), the study estimates that a well-designed cash-transfer program could reduce extreme poverty to under 1% of the global population and lift hundreds of millions of people — roughly 8% of the world’s population — out of life-threatening deprivation. Recipients would often still be economically disadvantaged by broader measures, but would no longer face immediate survival-level shortfalls.

Practical Considerations

Delivering cash at global scale would be an "enormous logistical undertaking," as Paul Niehaus, co-founder of GiveDirectly and a co-author, acknowledges. But the study’s premise is that it is feasible: digital payments, national registries, and AI-driven targeting can make transfers more accurate and harder to misdirect. The main obstacles are political — securing sustained funding, international coordination, and institutional commitment — rather than purely financial.

The geography of remaining extreme poverty matters: much of it now lies in sub-Saharan Africa, where growth has been slower and populations are growing rapidly. Without targeted interventions, the absolute number of people living in extreme poverty could rise despite global progress over the past three decades.

Why This Matters

The study reframes the debate: ending extreme poverty is not primarily a problem of affordability. The funds exist within global wealth and spending patterns; the challenge is mobilizing them, designing robust delivery systems, and overcoming political and institutional resistance. Whether through governments, philanthropists, or broad-based contributions, the authors argue the price is surprisingly small — and the moral case compelling.

Key actors mentioned: GiveDirectly (Paul Niehaus) and researchers including Joshua Blumenstock (UC Berkeley) — leaders in evidence-driven cash assistance research.