Survey of EA org leaders about what skills and experience they most need, their staff/​donations trade-offs, problem prioritisation, and more.

N.B. You may be in­ter­ested to read our fol­low-up post: Many EA orgs say they place a lot of fi­nan­cial value on their pre­vi­ous hire. What does that mean, if any­thing? And why aren’t they hiring faster?


For the third year run­ning we’ve sur­veyed lead­er­ship at EA or­gani­sa­tions about a range of is­sues where their views might be rele­vant to EAs’ ca­reer de­ci­sions:

What are the most press­ing tal­ent gaps in pro­fes­sional effec­tive al­tru­ism in 2018? And which prob­lems are most effec­tive to work on? New sur­vey of or­gani­sa­tional lead­ers.

It com­ple­ments the 2018 EA Sur­vey which aims to col­lect in­for­ma­tion about ev­ery­one who says “they can, how­ever loosely, be de­scribed as an effec­tive al­tru­ist.”

We asked lead­ers about:

  • what skills and ex­pe­rience they most need;

  • what skills and ex­pe­rience they think the com­mu­nity as a whole will need in the fu­ture;

  • how many dona­tions they’d be will­ing to forego for their lat­est hires;

  • their view on the rel­a­tive cost-effec­tive­ness of the differ­ent EA Funds, and which new funds they’d like to see;

  • how ur­gent their need for ex­tra dona­tions and staff is;

  • and var­i­ous other is­sues.

We also sur­veyed peo­ple who iden­tify as mem­bers of the EA com­mu­nity and work di­rectly on prob­lems like an­i­mal welfare and poverty, to see how their views on some of these ques­tions would com­pare.

Here are some of the find­ings:

  • EA or­gani­sa­tion lead­ers said ex­pe­rience with op­er­a­tions or man­age­ment, and gen­er­al­ist re­searchers are what their or­gani­sa­tions will need most of over the next five years.

  • They said the com­mu­nity as a whole will most need more gov­ern­ment and policy ex­perts, op­er­a­tions ex­pe­rience, ma­chine learn­ing/​AI tech­ni­cal ex­per­tise, and skil­led man­agers.

  • Most EA or­gani­sa­tions con­tinue to feel more ‘tal­ent con­strained’ than fund­ing con­strained, rat­ing them­selves as 2.8/​4 tal­ent con­strained and 1.5/​4 fund­ing con­strained.

  • Lead­ers thought the key bot­tle­neck for the com­mu­nity is to get More ded­i­cated peo­ple (e.g. work at EA orgs, re­search in AI safety/​biose­cu­rity/​eco­nomics, ETG over $1m) con­verted from mod­er­ate en­gage­ment. The sec­ond biggest is to in­crease im­pact of ex­ist­ing ded­i­cated peo­ple through e.g. bet­ter re­search, co­or­di­na­tion, de­ci­sion-mak­ing.

  • We asked lead­ers their views on the rel­a­tive cost-effec­tive­ness of dona­tions to four funds op­er­ated by the com­mu­nity. The me­dian view was that the Long-Term Fu­ture fund was twice as effec­tive as the EA Com­mu­nity fund, which in turn was 10 times more cost-effec­tive than the An­i­mal Welfare fund, and twenty times as cost-effec­tive as the Global Health and Devel­op­ment fund. In­di­vi­d­ual views on this ques­tion varied very widely, though 1828 re­spon­dents thought the Long-Term Fu­ture fund was the most effec­tive.

  • In ad­di­tion, we asked sev­eral com­mu­nity mem­bers work­ing di­rectly on an­i­mal welfare and global de­vel­op­ment for their views on the rel­a­tive cost-effec­tive­ness of dona­tions to these funds. About half these staff thought the fund in their own cause area was best, and about half thought ei­ther the EA Com­mu­nity fund or Long-Term Fu­ture fund was best. The me­dian re­spon­dent in that group thought that the An­i­mal Welfare fund was about 33% more cost-effec­tive than the Long-Term Fu­ture fund and the EA Com­mu­nity fund—which were rated equally cost-effec­tive—while the Global Health and Devel­op­ment fund was 33% as cost effec­tive as ei­ther of those two. How­ever, there was also a wide range of views among this group.

  • The or­gani­sa­tions sur­veyed were usu­ally will­ing to forego over a mil­lion dol­lars in ad­di­tional dona­tions to get the right per­son in a se­nior role 3 years ear­lier, or sev­eral hun­dred thou­sand dol­lars for a ju­nior hire.

Con­tinue read­ing for de­tails of the method and re­sults...

Most an­swers were similar to what we found in 2017, so next year we ex­pect to ei­ther ask differ­ent ques­tions or in­ter­view a smaller num­ber of peo­ple in greater depth and see whether their re­sponses change af­ter fur­ther re­flec­tion.