A charity’s annual report says it saved a life for every $4,000 spent last year. You give $4,000. Did you save a life?
Not necessarily. The report describes the average dollar. Your gift is the marginal dollar: the one that arrives after all the others. What it buys depends on what was left undone when it showed up.
Three cases
The program had a gap. It planned to reach 12 districts and could fund 10. Your dollars help fund the 11th. Your marginal result is close to the average, maybe better if the unfunded districts were also the highest-burden ones.
The program was full. Every planned district was funded. Your dollars go to reserves, to next year, or to a new activity the charity has not yet shown it can do well. Your marginal result is unknown and probably lower than the average.
The program was full and the charity expanded to use your money. The 13th district has less malaria than the first 12, because they were chosen in order of need. Same program, higher cost per life. This is the normal shape of things: the best opportunities get funded first, so the marginal dollar is almost always less cost-effective than the average one.
Why evaluators talk about “room for more funding”
GiveWell’s estimates, and its funding bar (8x its benchmark as of November 2025), are about marginal grants: what the next block of money to a specific program in a specific place will do. When a program’s best sites are funded, GiveWell either stops or reports a lower figure for the remaining sites. This is why its top charities’ cost-per-life figures move from year to year even when the programs do not change.
Charities that publish gaps are doing this work for you. LEEP said in late 2025 that its 2026 paint-program gap was filled and that new money would go to expansion and to spices and cosmetics. That is a charity telling you, correctly, that your marginal dollar now buys something different from its headline figure.
The mirror image: crowding
If a well-known program is oversubscribed, the marginal dollar can be worth nearly nothing this year. The money is not wasted; it is spent later or elsewhere. But a donor comparing two programs should compare their marginal values, and a famous charity can have a lower one than an obscure charity doing the same thing in a place nobody has funded.
This is why the 2025 aid cuts, which are terrible, also raised the value of a private donor’s marginal dollar in global health: programs that were full are now not.
How to act on it
- Ask the gap question. “How much more could you use this year and what would it do?” Any charity that can answer has thought about its marginal dollar.
- Prefer pooled funds when you cannot tell. GiveWell’s funds, ACE’s Recommended Charity Fund and the Founders Pledge climate and global health funds exist to move money to wherever the marginal value is highest that quarter. You cannot do that from outside.
- Do not chase last year’s number. A headline cost per outcome is a rear-view mirror. The evaluators’ current grant write-ups are the windshield.
- Give when gaps are open. Timing matters more than most donors think. Year-end giving is convenient; it is also when everyone else gives.
The one-sentence version
The question is never “how good is this charity?” It is “what will this charity do with the next dollar that it could not do without it?”
Method notes
Examples are stylized. GiveWell’s bar and LEEP’s funding statement are quoted from the sources above.
Sources
Figures come from the cited sources as of the dates shown. Cost-effectiveness estimates change as programs and evidence change; check the source before giving. We are not financial advisers and receive nothing from any charity named here.