Should you switch away from earning to give? Some considerations.

At EA Global, a com­mon dis­cus­sion topic was that di­rect work was be­com­ing in­creas­ingly higher pri­or­ity in com­par­i­son to earn­ing to give. I’ve helped to con­tribute to this con­ver­sa­tion, and I’m glad we’re talk­ing about this. This post is just to give some con­sid­er­a­tions on the other side, to en­sure that (i) we ap­prox­i­mate al­loca­tive effi­ciency, where the peo­ple who have the strongest com­par­a­tive ad­van­tage in earn­ing to give do so, and those who have strongest com­par­a­tive ad­van­tage in di­rect work do so; and that (ii) we don’t over­cor­rect. There were a few con­ver­sa­tions I had at EA Global where I thought there were some con­sid­er­a­tions that should be on the table that weren’t widely known about, and weren’t writ­ten up pub­li­cly, so I’ve cho­sen to write them up here. I should caveat that these are just my per­sonal off-the-cuff thoughts, rather than nec­es­sar­ily rep­re­sen­ta­tive of CEA or 80,000 Hours. This also isn’t meant to be a com­plete pic­ture; it’s just some con­sid­er­a­tions that I think might not cur­rently be widely known or fully un­der­stood.


The pri­mary rea­son peo­ple were less ex­cited by earn­ing to give was sim­ply that we’ve raised a lot of money already, and diminish­ing marginal re­turns means that ad­di­tional money be­comes com­par­a­tively less im­por­tant.

It’s true that, as a com­mu­nity, we’ve raised an awful lot of money; it’s very plau­si­ble that we’ve had com­par­a­tively greater suc­cess at mov­ing money to the high­est-pri­or­ity is­sues than con­vinc­ing peo­ple to work on those is­sues. What’s more, Open Philan­thropy has started mak­ing grants in the ar­eas that peo­ple in the com­mu­nity of­ten think of as high­est-pri­or­ity: fac­tory farm­ing, global catas­trophic risks, ar­tifi­cial in­tel­li­gence, and build­ing the effec­tive al­tru­ism com­mu­nity.

Th­ese facts make me be­lieve that fewer peo­ple should earn to give than I used to be­lieve sev­eral years ago. But I still think some peo­ple should earn to give; in my pre­vi­ous post a year ago I asked, “At this point in time, and on the mar­gin, what por­tion of al­tru­is­ti­cally mo­ti­vated grad­u­ates from a good uni­ver­sity, who are open to pur­su­ing any ca­reer path, should aim to earn to give in the long term?” The me­dian and mean an­swer among my­self, Ben Todd, Ro­man Duda and Rob Wiblin was 15%. This num­ber still seems about right to me: the peo­ple with the high­est com­par­a­tive ad­van­tage in earn­ing to give should con­tinue to earn to give; other peo­ple should do di­rect work (in­clud­ing re­search, ad­vo­cacy, policy, so­cially-mo­ti­vated en­trepreneur­ship, etc).

Alloca­tive Efficiency

My pri­mary con­cern go­ing for­ward is that we as a com­mu­nity fail on al­loca­tive effi­ciency. Differ­ent peo­ple vary con­sid­er­ably on both their dona­tion po­ten­tial and their po­ten­tial at di­rect work. Ideally, if 15% of peo­ple should earn to give long-term, then it’s the 15% of peo­ple who have the strongest com­par­a­tive ad­van­tage at earn­ing to give.

For this rea­son, I be­lieve that those peo­ple who have the strongest com­par­a­tive ad­van­tage at earn­ing to give, such as those work­ing in quan­ti­ta­tive trad­ing, or who are already far along in their earn­ing-to-give ca­reer, should at least wait and see over the next year or two how the com­mu­nity re­sponds and de­vel­ops be­fore switch­ing to di­rect work. To em­pha­sise, this is about com­par­a­tive ad­van­tage: if you have lots of amaz­ing op­tions, then the fact that you have an amaz­ing earn­ing to give op­tion doesn’t mean you should earn to give; for ex­am­ple, for some­one con­cerned about AI safety, who has the op­tion to work on AI safety at an AI lab, it’s plau­si­ble to me that they ought to do that even over a great earn­ing to give job in quan­ti­ta­tive trad­ing.**[later edit] For other peo­ple, for ex­am­ple peo­ple earn­ing to give in soft­ware en­g­ineer­ing, who have skills that would be use­ful in di­rect work, and who would be happy ei­ther earn­ing to give or do­ing di­rect work, I’d still en­courage them to se­ri­ously think about what sorts of di­rect work they could do.

In gen­eral, when con­sid­er­ing whether to do di­rect work or earn to give, you could ask your­self: am I in the top 15% of peo­ple in terms of com­par­a­tive ad­van­tage at earn­ing to give?


As well as en­sur­ing that we get al­loca­tive effi­ciency, it’s also pos­si­ble that the com­mu­nity will over­cor­rect to changes in cir­cum­stance. So, for peo­ple who are cur­rently pur­su­ing earn­ing to give and won­der­ing whether to switch to di­rect work, here are some coun­ter­vailing con­sid­er­a­tions to bear in mind:

  • You may differ from Open Phil or other ma­jor fun­ders in your as­sess­ment of the fund­ing gaps within what you think of as the high­est-pri­or­ity is­sues. As a hy­po­thet­i­cal: If you think that fac­tory farm­ing would still be the high­est-pri­or­ity prob­lem even if a fur­ther $100mn/​yr were put to­wards it, whereas Open Phil think that the room for more fund­ing is only $10mn/​yr, then earn­ing to give, from your per­spec­tive, would still be very valuable over the long-run, even though you’re both fund­ing it as much as you can right now.

  • Over time, your views or the views of Open Phil and other ma­jor fun­ders may change sub­stan­tially on what the top pri­ori­ties are. Even if you’re in perfect agree­ment right now, that might change in the fu­ture; earn­ing to give can be an im­por­tant hedge against this pos­si­bil­ity.

  • For some or­gani­sa­tions, Open Phil and other ma­jor fun­ders might wish to limit their dona­tion to a cer­tain % of their over­all bud­get, in or­der to en­sure that the or­gani­sa­tion doesn’t be­come overly de­pen­dent on a sin­gle donor. In some cir­cum­stances, this can ac­tu­ally in­crease the ar­gu­ment for earn­ing to give, be­cause ev­ery dol­lar you donate un­locks an ad­di­tional amount of room for more fund­ing from those %-capped ma­jor donors.

  • There will be giv­ing op­por­tu­ni­ties that Open Phil and other ma­jor fun­ders won’t look at, for ex­am­ple be­cause the fund­ing gap for the or­gani­sa­tion in ques­tion is com­par­a­tively small.

  • We can already see peo­ple in the com­mu­nity adapt­ing their plans on the ba­sis of the em­pha­sis away from earn­ing to give; so you need to take this ad­just­ment into ac­count. Even if you thought that too many peo­ple are earn­ing to give right now, this might not be true in two years, af­ter the ad­just­ment takes place.

  • Similarly, or­gani­sa­tions are able to in­crease their room for more fund­ing in re­sponse to a greater availa­bil­ity of fund­ing. For ex­am­ple, many or­gani­sa­tions that peo­ple in the com­mu­nity donate to pay sig­nifi­cantly less than mar­ket rates; it’s pos­si­ble that with a greater abun­dance of fund­ing they could pay more in or­der to at­tract more ex­pe­rienced peo­ple; or they could start hiring more ex­pert con­trac­tors, which are typ­i­cally ex­pen­sive; or they could come up with in­no­va­tive ways of spend­ing money that don’t re­quire hiring a lot of peo­ple.

  • Similarly, new or­gani­sa­tions within a par­tic­u­lar area may come into ex­is­tence if it’s widely known that there’s fund­ing for such or­gani­sa­tions.

  • In some cases, it’s pos­si­ble to ask the or­gani­sa­tions where you might work for the amount of dona­tions at which they’d be in­differ­ent be­tween hav­ing you work for them and hav­ing you donate more. This can be an awk­ward con­ver­sa­tion to have, but does en­able you to more di­rectly make a com­par­i­son be­tween earn­ing to give and di­rect work.

  • There may be fewer peo­ple earn­ing to give than you think. Only 10% of at­ten­dees at EAG were earn­ing to give as their long-term plan for im­pact. This is less than the 15% sug­ges­tion I made in my pre­vi­ous blog post on the topic. In an in­for­mal sur­vey of or­gani­sa­tions in the effec­tive al­tru­ism com­mu­nity done by 80,000 Hours at EA Global, re­spon­dents on av­er­age only claimed to be slightly more peo­ple-con­strained than fund­ing-con­strained.

The area I know in most depth is fund­ing of effec­tive al­tru­ism com­mu­nity-build­ing. In this area, all of the con­sid­er­a­tions above weigh on my mind; I think it would be a very pre­car­i­ous po­si­tion (both for in­surance and im­par­tial­ity rea­sons) if the EA com­mu­nity were heav­ily de­pen­dent on a sin­gle donor, or a small num­ber of donors. I wouldn’t be that sur­prised if in a few years we switched to em­pha­sis­ing how fund­ing-con­strained rather than peo­ple-con­strained we are (though I’m aware that other peo­ple dis­agree with me on the like­li­hood of this).

Ca­reer ad­vice is hard to give gen­eral recom­men­da­tions about, be­cause ev­ery­one’s cir­cum­stances and op­tions are so differ­ent. I’m glad that we’re now more heav­ily em­pha­sis­ing ca­reer paths other than earn­ing to give. But I think that if you’re par­tic­u­larly well-suited to earn­ing to give, com­pared to your other op­tions, or if you’d gain a lot of skills by earn­ing to give, or if you’d be par­tic­u­larly happy in that path, or if you are par­tic­u­larly un­sure about which prob­lems are high­est pri­or­ity to tackle, then it’s of­ten still a great op­tion for hav­ing a pos­i­tive im­pact, and you should be cau­tious about mov­ing on from that.

Thanks to Nick Beck­stead, Holden Karnofsky and Ben­jamin Todd for com­ments on an ear­lier draft.

**[Later Edit] To clar­ify again, this is about com­par­a­tive ad­van­tage, not ab­solute ad­van­tage: x has a com­par­a­tive ad­van­tage over y at pro­duc­ing G iff x can pro­duce G at a lower op­por­tu­nity cost than y. (In this post I’m most in­ter­ested in com­par­a­tive ad­van­tage within the EA com­mu­nity). Ex­am­ple: If Jane can earn to give and donate $100,000 per year, or do re­search and write 8 pa­pers per year; and Joe can earn to give and donate $50,000 per year or do re­search and write 3 pa­pers of the same qual­ity that Jane could write per year, then Jane has a com­par­a­tive ad­van­tage in re­search (giv­ing up 8 re­search pa­pers for ev­ery $100,000 donated) and Joe has a com­par­a­tive ad­van­tage at earn­ing to give (giv­ing up only 6 re­search pa­pers for ev­ery $100,000 donated).

[Se­cond later edit]: Dis­clo­sure: I am CEO of the Cen­tre of Effec­tive Altru­ism, which is funded in sig­nifi­cant part by the dona­tions of effec­tive al­tru­ists, in­clud­ing donors who earn to give. You can find out more about my back­ground at www.willi­am­