If you believe the most cost-effective organisation (at the margin) is over 100 times as cost-effective as your paid work, you can have more impact through donations than through paid work donating 1 % (= 1/100) of your gross income to the most cost-effective organisation, and more than 10 times as much impact through donations as through paid work donating 10 %. In this case, increasing your donations to the most cost-effectiveness organisation by x % would be 10 times as impactful as increasing the cost-effectiveness of your paid work by x %, which suggests focussing on donating more and better.

I think the most cost-effective organisation is over 100 times as cost-effective as the vast majority of impact-focussed organisations. My top candidates for the most cost-effective organisation are, ordered alphabetically, the Arthropoda Foundation, Shrimp Welfare Project (SWP), and Wild Animal Initiative (WAI). I believe any of these is:

• More than 100 times as cost-effective as the most cost-effective organisations in human welfare, including not only global health and development, but also global catastrophic risk. I estimate:
  • The Shrimp Welfare Project (SWP) has been:
    • 64.3 k times as cost-effective as GiveWell’s top charities (neglecting the effects of these on animals), which are often considered the best helping humans.
    • 19.6 k (= 64.3*10^3/3.28) times as cost-effective as Founders Pledge’s Climate Fund, which is often considered the best option to decrease the harm caused by global warming to humans.
    • 265 k (= 64.3*10^3*4.12) times as cost-effective as epidemic/pandemic preparedness.
    • 1.43 M times (= 64.3*10^3/0.045) as cost-effective as research and development (R&D) if this did not increase SWP’s funding, since Open Philanthropy (OP) estimated it is 4.5 % as cost-effective as GiveWell’s top charities. SWP received 1.82 M 2023-$ (= 1.47*10^6*1.24) during the year ended on 31 March 2024, which is 1.72*10^-8 (= 1.82*10^6/(106*10^12)) of the gross world product (GWP) in 2023, and OP estimated R&D has a benefit-to-cost ratio of 45. So I estimate SWP is 1.29 M (= 1/(1.72*10^-8)/45) times as cost-effective as R&D due to this increasing SWP’s funding.
  • Paying farmers to use more humane pesticides would be 23.7 k times as cost-effective as GiveWell’s top charities.
    • I guess that research on and advocacy for more humane pesticides would be way more cost-effective than paying farmers to use them more.
    • WAI does research on pesticides.
• Possibly more than 100 times as cost-effective as ones in animal welfare which are often considered to be among the most cost-effective. I estimate:
  • SWP has been 173 times as cost-effective as cage-free campaigns.
  • Paying farmers to use more humane pesticides would be 51.4 times as cost-effective as cage-free corporate campaigns helping hens, and guess that research on and advocacy for more humane pesticides would be way more cost-effective than paying farmers to use them more.

So I conclude increasing the impact of donations via donating more and better is the best strategy to maximise impact for the vast majority of people working in impact-focussed organisations.

You may disagree with my estimates which suggest prioritising animal welfare. However, I would say increasing the impact of donations is also the best strategy to maximise impact for (random) people working in the area they consider most cost-effective. Benjamin Todd thinks “it’s defensible to say that the best of all interventions in an area are about 10 times more effective than [as effective as] the mean, and perhaps as much as 100 times”, which is in agreement with the cost-effectiveness of interventions following a heavy-tailed distribution. If so, and jobs were uniformly distributed across interventions, a person in a random job within an area donating 10 % of their gross salary to the best interventions in the area could have 1 (= 0.1*10) to 10 (= 0.1*100) times as much impact through donations as through work. In reality, I assume there will be more jobs in less cost-effective interventions, as the best interventions only account for a small fraction of the overall funding. Based on Ben’s numbers, if there are 10 times as many people in jobs as cost-effective as a random one as in the most-effective jobs, a person in a random job within an area donating 10 % of their gross salary to the best interventions in the area would be 10 (= 1*10) to 100 (= 10*10) times as impactful as a person in the same job not donating.