Suppose you have $10,000 to give and no deadline. Give it today, or invest it at 5% real and give $43,000 in thirty years?
The case for waiting
It has three legs.
Compounding. Markets have historically returned more than the rate at which the cost of doing good rises. If the best program costs $4,000 per life today and $6,000 (in today’s dollars) in thirty years, your $43,000 then saves more lives than your $10,000 now.
Learning. We know more about what works every year. Money given in 2000 went to many programs later shown to be ineffective. Money held until the evidence is in avoids that.
Options. A future emergency, a technology that makes a problem suddenly cheap to solve, or a moment when a dollar has unusual leverage: cash on hand can take it; money already spent cannot.
Founders Pledge built a Patient Philanthropy Fund on this reasoning. It invests donations with the explicit plan of granting them decades from now, when they may buy more.
The case for now
Compounding runs on both sides. A child who survives malaria at two grows up, earns, and raises children. Health and income gains compound too, in ways that are hard to price but not zero. The comparison is not “return on stocks versus zero”; it is “return on stocks versus return on a saved life.”
The best opportunities may not wait. The cheapest life-saving programs exist because some countries are still poor and some diseases are still untreated. As countries develop, those opportunities disappear. The cost per life saved in the best programs has risen over the past two decades for exactly this reason. Waiting for it to fall is betting against the trend.
Learning cuts both ways. We already know a great deal. Four programs with cost per life saved in the $3,500 to $5,500 range are documented in detail. Waiting for better knowledge, when good knowledge exists, is a way of not deciding.
Money held is money at risk. Of your own change of heart, your heirs’ priorities, a market crash the year you meant to give, or a fund that drifts from its purpose.
Why 2026 tilts toward now
The 2025 collapse in government aid took the argument out of the seminar room. A Lancet Global Health model this February projected 9.4 million additional deaths by 2030 under current funding trends. Programs that were fully funded in 2024 have gaps in 2026. When the marginal dollar in the best programs buys more than it did two years ago, the expected return from waiting has to beat not only the market but a temporarily elevated return from giving.
Most people who take the patient argument seriously also concede this: patience is a strategy for money that has nowhere unusually good to go right now. That is not this year.
The same logic applies to the cheapest lever of all. The aid cuts were a policy choice, and the November 2026 elections decide who makes the next one. A vote costs nothing and cannot be invested for later; find out which races are on your ballot before you decide how much of your giving to hold back.
A split that respects both
- Give the impact share now, to programs with a measured cost per outcome and an open funding gap.
- If you want a patient allocation, make it explicit and small: a fixed slice into a fund designed for it, with a written purpose, rather than a growing balance in a donor-advised fund that “will get to it.”
- Revisit the split each year. The right answer depends on the gaps that year.
Method notes
The $43,000 figure is $10,000 compounded at 5% real for 30 years. Cost-per-life figures are GiveWell’s 2022 to 2024 averages. The claim that cost per life saved has risen over two decades reflects GiveWell’s published history of estimates for its top programs.
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.