A clarification/example from the 80,000 hours AI bot regarding diminishing marginal returns I found helpful:
“The key idea is that a charity’s average impact per dollar isn’t the same as what your next dollar achieves.
Concrete example: imagine a program distributing vitamin supplements in one region.
First $100k: reaches the villages closest to the road, with the worst deficiencies. Massive impact per dollar.
Next $500k: reaches more remote villages — more spent on transport, populations slightly less in need. Still good, but less per dollar.
Next $1M: the neediest people are already covered; now it’s trucking supplements to areas with milder deficiency at high logistical cost. Much weaker per dollar.
The charity’s average looks great — because the early dollars did so much. But your donation is the last dollar in, funding the marginal (weakest remaining) activity. That’s “diminishing marginal returns”: each additional dollar tends to buy less than the one before, because organisations naturally do the cheapest, highest-impact things first.
So the smart question isn’t “which charity has the best track record per dollar?” but “where will one more dollar do the most?” — which might be a smaller org that hasn’t yet exhausted its best opportunities, rather than a famous one that has.”
- And there will be more on marginal impact a few pages further in the handbook!
A clarification/example from the 80,000 hours AI bot regarding diminishing marginal returns I found helpful:
“The key idea is that a charity’s average impact per dollar isn’t the same as what your next dollar achieves.
Concrete example: imagine a program distributing vitamin supplements in one region.
First $100k: reaches the villages closest to the road, with the worst deficiencies. Massive impact per dollar.
Next $500k: reaches more remote villages — more spent on transport, populations slightly less in need. Still good, but less per dollar.
Next $1M: the neediest people are already covered; now it’s trucking supplements to areas with milder deficiency at high logistical cost. Much weaker per dollar.
The charity’s average looks great — because the early dollars did so much. But your donation is the last dollar in, funding the marginal (weakest remaining) activity. That’s “diminishing marginal returns”: each additional dollar tends to buy less than the one before, because organisations naturally do the cheapest, highest-impact things first.
So the smart question isn’t “which charity has the best track record per dollar?” but “where will one more dollar do the most?” — which might be a smaller org that hasn’t yet exhausted its best opportunities, rather than a famous one that has.”
- And there will be more on marginal impact a few pages further in the handbook!