A new study by U.S.-based academics has found that ending global poverty could be “surprisingly affordable,” requiring about 0.3% — or $318 billion — of global gross domestic product (GDP) to reduce the worldwide poverty rate from 10% to 1%.
The findings were released in a Dec. 15 news announcement by the Center for Effective Global Action (CEGA), a research, training, and innovation hub headquartered at the University of California, Berkeley, and UC San Diego, and co-authored by researchers from Stanford University.
The report "What Would It Cost to End Extreme Poverty?" was originally published on Dec. 9, 2025.
“The share of the world’s population living in extreme poverty—defined as living on less than $2.15 in 2017 PPP dollars per day—has declined dramatically, from an estimated 41% in 1981 to 8% in 2024,” said the researchers in the study report.
“This progress has historically taken place in conjunction with economic growth.”
However, this trajectory may not continue, the researchers opined, saying that analyzing current growth trends, it can be concluded that progress against extreme poverty will come to a halt.
“And in either case the fact that hundreds of millions of people live below even very low consumption thresholds today raises an immediate policy question: could extreme poverty be eliminated more rapidly through redistribution?”
While current proposals suggest ending extreme poverty is financially manageable at just 0.08 percent of global GDP, implementing these targeted transfers is currently impossible because existing survey data is too limited to identify the specific needs of most households, the report said.
Even so, the study findings show that advances in state-of-the-art statistical learning methods —dubbed “old data, new methods”— can help “revitalize the global fight against extreme poverty,” according to CEGA, by using advanced “gap-minimising” statistical models to target aid, ensuring that ending poverty is 81% cheaper than implementing a Universal Basic Income.
Finding positive solutions to the issue can therefore move beyond hypotheses to real-world benchmarks and decisions. This approach is supported by the idea of harnessing artificial intelligence “to address pressing social issues.”
“[The study] estimates that simply providing direct income transfers [to each household] sufficient to lift almost everyone above the $2.15/day poverty line would cost just 0.3% of global GDP,” said CEGA. “This offers one simple and surprisingly affordable path toward eradicating poverty."
By comparison, the figures given to eradicate global poverty equate to about 50% more than recent expenditure on foreign aid but only a seventh as much as the world spends on alcohol each year.
In other words, the cost equates to only a third of 1% of global GDP: “somewhat more than was previously spent on foreign aid (0.21%), but a fraction of what the world already spends on alcohol (2.2%) or cosmetics (0.6%).”
The study also showed that a universal basic income (UBI) set at the extreme poverty line would cost about five times the cost of the data-driven approach. In most countries, data-driven targeting can thus reduce poverty by the same amount as UBI at a fraction of the cost, said CEGA.
“The numbers tell us it isn’t crazy to set our sights on a big, ambitious goal,” said Paul Niehaus, professor of Economics at UC San Diego and co-author of the study.
“That’s partly because extreme poverty has already fallen so much in recent decades, alongside economic growth, but also because advances in data science have made it possible to design shovel-ready policies that get the most help to the poorest people.”
Furthermore, another comparable for eradicating global poverty is that if every American earning the median U.S. personal income of $45,000 per year sacrificed 0.3% of their salary this would equate to a personal contribution of $135 per year.
“With today’s levels of private wealth, over one hundred individuals could end poverty in an entire country by committing the earnings on their assets,” added CEGA, providing another comparison.
The study was conducted by Roshni Sahoo (Stanford University), Joshua Blumenstock (UC Berkeley), Paul Niehaus (UC San Diego), Leo Selker (UC Berkeley), and Stefan Wager (Stanford University). Their work utilised an integration of economics, statistical learning, and global development.
