“Imagine what we could do for the poor if we just spent NASA’s budget on them instead” is one of the most reliable lines on the internet, trotted out every time a rocket launches or a budget bill passes. It sounds like arithmetic. It is not arithmetic. Run the actual numbers against a real poverty-economics study and the comparison collapses almost immediately, not because ending poverty is too expensive to take seriously, but because NASA’s budget is nowhere near large enough to be the thing standing in the way of it.
Congress finalized NASA’s fiscal year 2026 budget at $24.4 billion in January 2026, rejecting a White House proposal that would have cut the agency closer to $18.8 billion. It’s a real number, and a real fight over it played out in Washington. But set next to what a new working paper says it would actually cost to end extreme poverty worldwide, $24.4 billion isn’t the number that settles the argument. It’s the number that shows how far short the argument falls.
What ending extreme poverty actually costs
A December 2025 National Bureau of Economic Research working paper, led by researchers at UC Berkeley, Stanford and UC San Diego, set out to answer the cost question with more precision than it usually gets. Using household survey data from 23 low-income countries and machine-learning methods to identify exactly who needs how much, the team calculated that lifting the extreme poor in those 23 countries to $2.15 a day, the benchmark used at the time of the study, would cost roughly $170 billion a year if the money were optimally targeted rather than spread evenly. (The World Bank has since raised its official extreme-poverty line to $3 a day — a change the researchers note would push the true cost above $318 billion, not below it.)
Extrapolated across the full 85 countries the researchers estimate account for the world’s extreme poor, some 533 million people, that figure rises to $318 billion annually. That’s the number worth sitting with: not $170 billion, which covers a partial slice of the problem, but $318 billion, which is what actually ending the thing looks like at global scale. As one of the study’s authors, Joshua Blumenstock of UC Berkeley’s School of Information, put it, the figure represents “0.3% of global GDP,” a share he compares directly to the roughly 2.2% of global GDP the world spends every year on alcohol.
Where the popular version of the math breaks
Line the two numbers up and the arithmetic behind “fund poverty instead of space” falls apart almost immediately. NASA’s FY2026 appropriations bill — the funding Congress just fought over — covers less than 8% of what the researchers say is needed to end extreme poverty for a single year (NASA’s total FY26 funding, once a separate $10 billion allocation from other legislation is included, comes to about $27.5 billion — still under 9% of the poverty figure.) Even if every dollar NASA spends were redirected tomorrow, with zero loss to overhead or delivery costs, the money would run out having addressed roughly one thirteenth of the problem it was supposedly going to solve.
That’s the part the internet version of this argument tends to skip. It isn’t just that redirecting NASA’s budget would be a political long shot. It’s that the number being redirected was never close to sufficient in the first place. The comparison implicitly treats NASA’s budget as though it’s in the same weight class as global poverty spending, when it’s closer to one-thirteenth the size. A single national space agency’s science and exploration budget, compared against a genuinely global economic problem spanning 85 countries and more than half a billion people, was always going to look small by comparison. It just turns out to be smaller than most people assume when they reach for it as an example.
The actual scale of the number, and why it’s smaller than it sounds
None of this means ending extreme poverty is out of reach, and that’s the part of the study that tends to get lost once the NASA comparison enters the conversation. $318 billion sounds enormous stated on its own. Stated as a share of the global economy, 0.3%, it sits below what the world spends annually on cosmetics (0.6% of global GDP, by Blumenstock’s own comparison) and far below global alcohol spending. Roshni Sahoo, the study’s lead author and a Stanford doctoral researcher, built the underlying model as what she has described as a statistical targeting problem: identify precisely who is poorest, using real survey data across 23 countries, and allocate a fixed budget to close the gap as efficiently as possible rather than spreading it thin.
That targeting detail matters, because the naive version of ending poverty, an equal transfer to every poor household regardless of how far below the line they sit, would cost dramatically more than either figure the researchers arrived at. The $170 billion and $318 billion numbers both assume the money goes to the right people in the right amounts, which is a genuinely hard logistical and political problem even once the dollar figure is settled. Money alone doesn’t distribute itself.
What the comparison is actually good for
The honest version of this comparison isn’t “cancel the space program,” and it isn’t “poverty is unsolvable.” It’s that global problems and national agency budgets are different units of measurement, and collapsing them into the same sentence produces a satisfying line that doesn’t survive contact with either number. NASA’s $24.4 billion is a meaningful sum inside one country’s federal budget. $318 billion is a meaningful sum inside the global economy, a genuinely small share of world output for a problem this size, but a share nearly thirteen times larger than what any single space agency has to give. The math behind ending extreme poverty turns out to be more encouraging than the popular framing suggests. It just isn’t sitting inside NASA’s checking account, and it never was.