Ending extreme poverty costs about a hundred dollars per person per year. That number is smaller than what is already being spent on aid, and this volume is obliged to sit with what follows.
Seven hundred million people live below the World Bank's extreme-poverty line of $2.15 a day. UNU-WIDER put the cost of closing that income gap at $70 billion a year in 2023 dollars — which they note is one tenth of one percent of the OECD high-income countries' gross national income.
A hundred dollars a person a year. That is the whole of it, at that line, and it is the figure this chapter is about.
Read the third and fourth rows again, because they are the ones that matter to this volume.
The money is not merely findable somewhere. It is already being spent. Official development assistance ran to $224 billion in 2023 — $193 billion once you remove the $31 billion spent inside donor countries on refugees — which is nearly three times the figure. Add the $33 billion a year that IFPRI and Ceres2030 estimate would end hunger sustainably rather than by transfer, and the pair still comes to about half of existing net aid.
So this is not a financing problem, and therefore it cannot be solved by finding money. That sentence is the reason the chapter is in Book X rather than in a book about economics. It is a custody question: the resource exists, is held, moves in roughly the right quantity, and does not arrive.
A. We do not have the resources.
Refuted above. This is the only one of the three that arithmetic settles, and it settles it against. It is also the answer people reach for first, which is why the chapter spends its first half on a table.
B. We have them, and the transfer is the problem.
Moyo's argument, and the subject of ch03 of this volume: sustained government-to-government transfers corrode the institutions they pass through, so more of the same makes it worse. On this account the $70 billion is available, has largely been spent several times over, and spending it again is not the move.
C. We have them, and nobody moves first.
A coordination account. The low-action state is a stable equilibrium and remains stable however much accumulates beside it — modelled in WP-126, where the stuck equilibrium sits under 1.2% of actors acting, and the intervention that would tip it is under a fifth of the programme.
B says the flow is the damage and should be cut. C says the flow is too small and too uncoordinated and should be tipped. They cannot both be the main story.
This chapter does not adjudicate. That is honest and it is also unfinished — the evidence that would separate them exists and reading it is the next piece of work, not a gap that argument can close.
What the chapter does refuse is letting A stand in for either. A is the answer that gets given, and it is the one that is false.
Not the cost of ending poverty. It is the cost of closing an income gap to a line, by transfer, for one year, assuming perfect targeting. UNU-WIDER say so themselves; delivery and administration sit on top, and real programmes reach the wrong people and miss the right ones.
Not a one-off. It is per year, and it buys the line being crossed, not the capacity to stay across it. That capacity is the separate $33 billion, and that one is investment rather than transfer — which is also why the two numbers should never be added without saying which is which.
Not applicable to famine. Acute famine today — Sudan, Gaza, Yemen — is conflict famine, and its binding constraint is access, not money. No transfer reaches a besieged population. A chapter that let a poverty figure stand in for a famine figure would be making exactly the category error it accuses others of in §3, and the two problems are not the same problem.
Not a plan. It is a price.
The distance between a price and a plan is where this entire volume lives. The Ishango bone is in Brussels not because anyone could not afford to move it. Receiving capacity turned out to exist and be measured, and the constraint turned out to be elsewhere. Each time, the resource was present and the arrival was not, and each time the thing worth writing down was the gap between them.
Six gaps in the script. The one that is genuinely unfinished rather than merely declared: B versus C is not adjudicated. Two incompatible accounts of the same fact, both serious, pointing opposite ways on what to do. The chapter sets them side by side and stops, and that is the correct place to stop today and the wrong place to stop permanently.
And every figure here is cited, not measured. The arithmetic is checked; the inputs are taken on the authority of UNU-WIDER, the OECD, IFPRI and SIPRI, and this chapter is worth exactly what they are.
Producing script: book10/ch06-the-price-of-the-thing-verify.py — 4 sections, 6 gaps. CITED: UNU-WIDER, new estimates of the cost of ending poverty; IFPRI / Ceres2030; SIPRI, world military expenditure 2025; OECD ODA 2023 via UNU-WIDER. Related: ch03, ch04, WP-126, WP-41.