You might say “well there’s a 50-percentage-point difference at each of those two steps” and think I’m being inconsistent in donating to AMF and not GAVI. But if I try some expectation-value-type calculation, I’ll be multiplying the impact of AMF’s work by 50% and getting something comparable to SCI, but getting something close to zero for GAVI.
But SCI is far more likely to need the money soon than AMF.
Probability that they’ll need my money soon:
GAVI: ~0%
AMF: ~50%
SCI: ~100%
You might say “well there’s a 50-percentage-point difference at each of those two steps” and think I’m being inconsistent in donating to AMF and not GAVI. But if I try some expectation-value-type calculation, I’ll be multiplying the impact of AMF’s work by 50% and getting something comparable to SCI, but getting something close to zero for GAVI.