A Framework for Thinking about the EA Labor Market

Note: Views mine, not my em­ployer’s. Thanks to David Re­in­stein, Peter Hur­ford, and oth­ers who pro­vided helpful feed­back; any er­rors are mine alone.


Since 2015, the EA com­mu­nity has in­creas­ingly dis­cussed tal­ent con­straints. That dis­cus­sion gen­er­ally hasn’t taken place in the lan­guage of­ten used to dis­cuss la­bor mar­kets, that of la­bor eco­nomics. I ar­gue that stan­dard frame­works like la­bor sup­ply and de­mand mod­els can provide con­sid­er­able in­sight into why EA or­ga­ni­za­tions and the broader EA com­mu­nity ex­pe­rience short­ages (and sur­pluses) of var­i­ous skills and how to re­solve these im­bal­ances.

Through an eco­nomics lens, or­ga­ni­za­tions that offer non-com­pet­i­tive com­pen­sa­tion should gen­er­ally ex­pect to find it hard to hire the best can­di­dates. It’s well known that the non­profit sec­tor as a whole pays sig­nifi­cantly less than the for-profit sec­tor, but there isn’t much hard data about how com­pen­sa­tion at EA or­ga­ni­za­tions com­pares to what differ­ent types of can­di­dates could al­ter­na­tively earn. I pro­pose sur­vey­ing EA or­ga­ni­za­tions to get ac­tion­able in­for­ma­tion about their com­pet­i­tive­ness, and us­ing that data to guide fu­ture de­ci­sions.

Defin­ing Ta­lent Constraints

When we talk about “tal­ent con­straints” in EA, what do we mean by the term?

80K has offered a work­ing defi­ni­tion of “tal­ent con­strained”:

“An or­ga­ni­za­tion is tal­ent con­strained when, for some­one who could take (a rea­son­ably im­por­tant) job at that or­ga­ni­za­tion, they would typ­i­cally con­tribute more to that or­ga­ni­za­tion by tak­ing the job than earn­ing to give.”

They go on to note that this defi­ni­tion is im­pre­cise and some­what prob­le­matic:

“While we think this fram­ing can some­times be use­ful, it also has some prob­lems. For ex­am­ple, this defi­ni­tion seems less use­ful when an or­ga­ni­za­tion’s best po­ten­tial hires don’t have very high earn­ing po­ten­tial and wouldn’t be very good fun­ders.”

This im­pre­ci­sion has the po­ten­tial to cre­ate am­bi­guity and con­fu­sion. I pro­pose an al­ter­na­tive defi­ni­tion:

An or­ga­ni­za­tion is tal­ent con­strained when it doesn’t have (and/​or can’t hire rea­son­ably eas­ily) the peo­ple it needs de­spite offer­ing com­pet­i­tive com­pen­sa­tion.

This defi­ni­tion (which I’ll use for the re­main­der of this post) may not be perfect[1], but it has the ad­van­tages of be­ing sim­ple and ap­pli­ca­ble across or­ga­ni­za­tions, causes, roles, and ge­ogra­phies. Per­haps most im­por­tantly, it’s an in­tu­itive defi­ni­tion: if some­one hears the term “tal­ent con­strained” it will likely con­jure up images of an or­ga­ni­za­tion that doesn’t have the tal­ent it needs, rather than an or­ga­ni­za­tion where (some) po­ten­tial sup­port­ers would be more valuable as em­ploy­ees than donors.

The defi­ni­tion also works in de­scribing more spe­cific skill gaps if one sim­ply states ex­plic­itly which skills one is refer­ring to. In gen­eral, I sug­gest peo­ple fol­low 80K’s ad­vice re­gard­ing speci­fic­ity: “It’s nearly always clearer to talk about the spe­cific needs… ideally down to the level of spe­cific pro­files of peo­ple, rather than tal­ent… in gen­eral.”[2]

The im­por­tance of com­pet­i­tive compensation

Through the lens of la­bor eco­nomics, which “looks at the sup­pli­ers of la­bor ser­vices (work­ers) and the de­man­ders of la­bor ser­vices (em­ploy­ers), and at­tempts to un­der­stand the re­sult­ing pat­tern of wages, em­ploy­ment, and in­come”, a tal­ent con­strained or­ga­ni­za­tion should con­sider rais­ing salaries.[3] When wages are struc­turally sup­pressed, through mechanisms like wage ceilings that limit how much peo­ple are paid, la­bor short­ages gen­er­ally re­sult.

It’s widely un­der­stood that the non­profit sec­tor pays sig­nifi­cantly less over­all than the for-profit sec­tor, and “much of the liter­a­ture on pay in non­prof­its” ad­dresses this dis­crep­ancy. This means that non­prof­its rou­tinely lose out on the tal­ent they need be­cause their salaries can’t com­pete with for-profit firms, as Dan Pal­lotta ar­gued in his 2013 TED Talk:

“The me­dian com­pen­sa­tion for a Stan­ford MBA, with bonus, at the age of 38, was 400,000 dol­lars. Mean­while, for the same year, the av­er­age salary for the CEO of a $5 mil­lion-plus med­i­cal char­ity in the U.S. was 232,000 dol­lars, and for a hunger char­ity, 84,000 dol­lars. Now, there’s no way you’re go­ing to get a lot of peo­ple with $400,000 tal­ent to make a $316,000 sac­ri­fice ev­ery year to be­come the CEO of a hunger char­ity.”

Ex­ac­er­bat­ing the prob­lem, while for-profit com­pa­nies rou­tinely re­ward the em­ploy­ees who have the largest im­pact on the bot­tom line, non­profit com­pen­sa­tion rarely in­cen­tivizes em­ploy­ees to achieve more so­cial im­pact. An anal­y­sis of the fi­nan­cial records of over 27,000 US non­prof­its found “the pay of non­profit chief ex­ec­u­tive officers (CEOs) is strongly pred­i­cated on what man­agers in similar-sized or­ga­ni­za­tions re­ceive… [and] that non­profit ex­ec­u­tive com­pen­sa­tion is only mod­estly af­fected by CEO perfor­mance, as mea­sured ei­ther by im­proved fund-rais­ing re­sults or bet­ter ad­minis­tra­tive effi­ciency.” Th­ese dy­nam­ics sug­gests there could be a large pool of tal­ented work­ers who would be will­ing to work in the non­profit sec­tor if the com­pen­sa­tion weren’t so low and so mis­al­igned with perfor­mance.

Turn­ing to an ex­am­ple from the EA com­mu­nity (where I sus­pect or­ga­ni­za­tions gen­er­ally pay less than what com­pa­rable non­prof­its do)[4], The Cen­tre for Effec­tive Altru­ism is hiring a new CEO. Should it re­strict its search to can­di­dates will­ing to show their com­mit­ment by pledg­ing ev­ery­thing they earn above a mod­est amount to effec­tive char­i­ties? (Purely hy­po­thet­i­cal ques­tion, I have no rea­son to think CEA is do­ing this).

That ap­proach would definitely pro­duce mis­sion al­igned can­di­dates. But I think it would be fool­ish given the ex­traor­di­nary im­por­tance of get­ting the best pos­si­ble per­son in that high-lev­er­age po­si­tion. What if that per­son won’t (or can’t) make the kind of sac­ri­fice Pal­lotta de­scribes? Is it worth los­ing out on the enor­mous in­cre­men­tal im­pact the per­son could pro­duce?

How com­pet­i­tive are EA or­ga­ni­za­tions?

To gauge whether an EA or­ga­ni­za­tion is com­pet­i­tive or not, we can bench­mark its salaries against what can­di­dates could earn on the open mar­ket, at other non­prof­its, or at other EA or­ga­ni­za­tions. Th­ese com­par­i­sons help define a can­di­date’s op­por­tu­nity cost (i.e. what they would forgo by work­ing in the EA ecosys­tem in­stead of pur­su­ing other op­por­tu­ni­ties). And, as the fol­low­ing sim­plified flowchart shows, they can also help define why tal­ent con­straints ex­ist and what can be done about them. (If you’re not sure how to fol­low the flow chart, there is a guided ex­am­ple in this foot­note).[5]

EA la­bor mar­ket supply

In­creas­ing com­pen­sa­tion can of­ten help at­tract tal­ent.[6] But I be­lieve many EAs are re­sis­tant to rais­ing salaries as a way to close tal­ent gaps be­cause they con­flate will­ing­ness to work for low pay with fit for a job. One re­spon­dent to the 2017 EA Lead­ers Sur­vey was ex­plicit about this: “Rais­ing salaries… is un­likely to be very helpful for at­tract­ing top tal­ent since the most suit­able can­di­dates are also the most al­tru­is­tic ones.”

While the “most al­tru­is­tic” can­di­dates might be the “most suit­able” all else be­ing equal, for prac­ti­cal pur­poses all else is never equal. As Holden Karnofsky ob­served about GiveWell’s ex­pe­rience, “it’s cer­tainly the case that some peo­ple re­quire sub­stan­tially more pay than oth­ers (based on differ­ent ca­reer stages, etc.) even when they buy heav­ily into the mis­sion.” In other words, will­ing­ness to work for be­low mar­ket rate is a poor proxy for both a can­di­date’s al­tru­ism and their job fit.

Ta­lent con­straints get ex­ac­er­bated when we (in Pal­lotta’s words) “con­fuse moral­ity with fru­gal­ity”. La­bor sup­ply curves shift in­ward, re­duc­ing the quan­tity of la­bor sup­plied, when can­di­dates have rel­a­tively higher earn­ing power in other in­dus­tries and when bar­ri­ers to en­try are in place (like cul­tural ex­pec­ta­tions of deeply dis­counted wage scales).[7]

Lower salaries don’t just re­duce the over­all sup­ply of la­bor, they also skew which types of can­di­dates end up work­ing, or not work­ing, in EA roles. EA lead­ers who were sur­veyed about tal­ent in 2018 think the com­mu­nity’s biggest skill gaps are in ar­eas that re­quire sig­nifi­cant ex­pe­rience like ex­per­tise in gov­ern­ment, policy, or AI, or skills in op­er­a­tions or man­age­ment. Th­ese la­bor short­ages are in sharp con­trast to the job mar­ket for EA roles that re­quire less ex­pe­rience, where even highly cre­den­tialed and mis­sion al­igned young can­di­dates are find­ing that it is re­ally re­ally hard to get hired by an EA or­ga­ni­za­tion.

The EA com­mu­nity’s youth­ful de­mo­graph­ics no doubt ex­plain much of this dis­crep­ancy, and re­lated fac­tors like net­work and founder effects likely play sig­nifi­cant roles as well. Th­ese is­sues make it all the more im­por­tant that EA or­ga­ni­za­tions re­con­sider other prac­tices like low salaries that ex­ac­er­bate these im­bal­ances.[8]

Low salaries make it rel­a­tively harder to find ex­pe­rienced can­di­dates than in­ex­pe­rienced ones be­cause of sev­eral fac­tors that shift the la­bor sup­ply curves on a rel­a­tive ba­sis:

Earn­ing power in al­ter­na­tive jobs. Ex­pe­rienced can­di­dates gen­er­ally have bet­ter pay­ing al­ter­na­tives than their in­ex­pe­rienced coun­ter­parts (i.e. they have higher op­por­tu­nity costs). Ex­pe­rienced can­di­dates will of­ten be sac­ri­fic­ing hun­dreds of thou­sands of dol­lars or more to work for an EA or­ga­ni­za­tion; that would be very rare for ju­nior can­di­dates.

Bar­ri­ers to en­try: Ex­pe­rienced can­di­dates are more likely to have de­pen­dents and mort­gages, and for them work­ing at an EA or­ga­ni­za­tion is more likely to in­volve a psy­cholog­i­cally difficult large pay cut. In­ex­pe­rienced can­di­dates, on the other hand, may see work­ing in EA as the path of least re­sis­tance.

Non-mon­e­tary com­pen­sa­tion. As Milan Griffes has ar­gued, EA jobs provide scarce non-mon­e­tary goods like “so­cial sta­tus, life-ori­en­ta­tion, a sense of hav­ing near-max­i­mal im­pact, and be­ing part of a value-al­igned, elite tribe.” Younger can­di­dates likely per­ceive more value from these par­tic­u­lar fac­tors. (More ex­pe­rienced can­di­dates, by con­trast, would likely per­ceive rel­a­tively more value on non-mon­e­tary com­pen­sa­tion in the form of flex­ible work­ing hours or paid parental leave).

Learn­ing more about EA’s tal­ent gaps and how to solve them

As 80K puts it, “Skill bot­tle­necks are a mat­ter of de­gree” and these de­grees can vary sig­nifi­cantly de­pend­ing on the spe­cific skills in ques­tion. But we have lit­tle hard data to quan­tify skill gaps across var­i­ous ar­eas.

I sug­gest adding new ques­tions to the next tal­ent sur­vey of EA or­ga­ni­za­tions to help cap­ture some of these nu­ances.[9] Th­ese ques­tions will hope­fully make it eas­ier to an­swer ques­tions like: Are EA or­ga­ni­za­tions tal­ent con­strained? If so, which sorts of or­ga­ni­za­tions and which sorts of tal­ent? How large are these con­straints? What can be done about them?

New ques­tions (which should con­tinue past sur­veys’ prac­tice of ask­ing the same ques­tions about both a ju­nior and se­nior hire and anonymiz­ing or­ga­ni­za­tional re­sponses due to the sen­si­tivity of the in­for­ma­tion in­volved):

● Gen­er­ally speak­ing, how easy/​difficult do you cur­rently find it to fill roles at your or­ga­ni­za­tion? (Scale of 1 = Very difficult to 5 = Very easy)

● What do you pay cur­rent em­ploy­ees rel­a­tive to what they could earn on open mar­ket, in­clud­ing jobs in the for-profit sec­tor? (Mul­ti­ple choice: More; about the same; 0-10% less; 11-20% less, etc)

● What do you pay cur­rent em­ploy­ees rel­a­tive to what they could oth­er­wise earn in the non­profit sec­tor (in­clud­ing all non­prof­its not just EA or­ga­ni­za­tions)? (Mul­ti­ple choice: More; about the same; 0-10% less; 11-20% less, etc)

● What do you pay cur­rent em­ploy­ees rel­a­tive to what they could oth­er­wise earn at an­other EA or­ga­ni­za­tion? (Mul­ti­ple choice: More; about the same; 0-10% less; 11-20% less, etc)

● How much do you plan to offer fu­ture hires rel­a­tive to cur­rent em­ploy­ees in similar roles? (Mul­ti­ple choice: More; Less; About the same. If some­one an­swers “more”: Do you plan to in­crease salaries for ex­ist­ing em­ploy­ees? What fac­tors into this de­ci­sion?[10]

● Would any of the fol­low­ing steps be helpful in clos­ing your or­ga­ni­za­tion’s tal­ent gaps? (Rate the fol­low­ing op­tions on a Scale of 1 =Not at all helpful to 5 = Ex­tremely helpful):

○ In­creas­ing salaries

○ In­vest­ing in recruiting

○ In­creased pub­lic­ity of position

○ Bet­ter ac­cess to in­ter­ested candidates

○ Other (please de­scribe)

This in­for­ma­tion is rel­a­tively easy to col­lect and in­ter­pret, cor­re­sponds to a real-world de­ci­sion or­ga­ni­za­tions make (how much to pay), is grounded in ob­serv­able data (mar­ket wages), and is com­pa­rable across roles, or­ga­ni­za­tions, ge­ogra­phies, and causes.[11] To miti­gate the cost of data col­lec­tion, I sug­gest aban­don­ing sur­vey ques­tions about non-tra­di­tional la­bor met­rics that seem to be pro­duc­ing noisy data and to gen­er­ally be caus­ing con­fu­sion.[12] If we col­lect com­pen­sa­tion data and re­spon­dents in­di­cate which types of roles they’re think­ing about as ju­nior and se­nior (which will vary across or­ga­ni­za­tions), we’ll have a rich and ac­tion­able pic­ture how EA com­pen­sa­tion stacks up to the com­pe­ti­tion for var­i­ous types of skills.


My hunch is that for the most part tal­ent con­straints are, and will con­tinue to be, com­mon among EA or­ga­ni­za­tions. The world’s biggest prob­lems re­quire sig­nifi­cant and di­verse tal­ent, es­pe­cially as ex­ist­ing or­ga­ni­za­tions con­tinue to grow and new ones emerge. And if EA or­ga­ni­za­tions offer com­pen­sa­tion that is as non-com­pet­i­tive as most non­prof­its, they should ex­pect chronic strug­gles at­tract­ing the can­di­dates they need (most no­tably in roles re­quiring spe­cial­ized ex­per­tise).

To solve this prob­lem, I think EA or­ga­ni­za­tions need to ap­proach com­pen­sa­tion in a way that’s ori­ented around im­pact rather than con­ven­tion, even if it some­times means pay­ing more than most non­prof­its would for a role. I sus­pect it would be very helpful to strate­gi­cally and op­por­tunis­ti­cally in­crease com­pen­sa­tion to try to tap a vein of skil­led can­di­dates that are will­ing to work in the non­profit sec­tor for a dis­count, but not as big a dis­count as most non­prof­its might ex­pect. [13] This might re­quire a shift in mind­set from view­ing tal­ent gaps and fund­ing gaps as op­po­si­tional to see­ing the lat­ter as a fre­quent cause of the former.[14] In­creas­ing com­pen­sa­tion won’t be a cure all (for an ex­cel­lent overview of why there can some­times be prac­ti­cal difficul­ties in trad­ing money for ca­pac­ity, see GiveWell’s 2013 blog post We can’t (sim­ply) buy ca­pac­ity), but my hunch is it would be a helpful first step.

How­ever, there’s no need for the EA com­mu­nity to rely on my in­tu­ition or any­one else’s when there’s hard data we can use to guide our de­ci­sions. I pro­pose sur­vey­ing EA em­ploy­ers to see how they pay rel­a­tive to the non­prof­its and for-prof­its they com­pete with for tal­ent. This would rad­i­cally im­prove our un­der­stand­ing of how much tal­ent EA or­ga­ni­za­tions want at differ­ent price lev­els; in eco­nomics terms, we’d be learn­ing what the la­bor de­mand curve looks like. I also en­thu­si­as­ti­cally sup­port pro­pos­als to use the EA sur­vey to learn more about EA job seek­ers, which would im­prove our un­der­stand­ing of the cor­re­spond­ing la­bor sup­ply curve.

If we think about the EA la­bor mar­ket in terms of la­bor sup­ply and de­mand, we’ll be able to lev­er­age the efforts of many very smart peo­ple who have already thought about a lot of the is­sues we care about.[15] They can help us make in­formed de­ci­sions about cru­cial ques­tions that will ul­ti­mately shape much of the EA la­bor mar­ket: How much should EA or­ga­ni­za­tions com­pete with the for-profit sec­tor for top tal­ent? How can they fund the tal­ent they need (and other op­er­a­tional ex­penses)? And if they can’t or won’t be­come more com­pet­i­tive with for-profit firms, how do EA or­ga­ni­za­tions ex­pect to find the tal­ent they need to achieve their am­bi­tious goals?

  1. I wel­come thoughts on how to bet­ter cal­ibrate and op­er­a­tional­ize “can’t hire rea­son­ably eas­ily” in the most mean­ingful way and/​or ideas for al­ter­na­tive defi­ni­tions. ↩︎

  2. While I dis­agree with 80K’s defi­ni­tion of tal­ent con­straints and some of their other think­ing on the is­sue, I want to em­pha­size that this con­sti­tutes a very minor cri­tique. 80K has done an ex­cel­lent job draw­ing at­ten­tion to the im­por­tance of the EA la­bor mar­ket, and has helped a lot of peo­ple achieve much more im­pact with their hu­man cap­i­tal than they oth­er­wise would. 80K also de­serves ap­plause for rec­og­niz­ing that some of its con­tent was caus­ing con­fu­sion and pub­lish­ing a clar­ifi­ca­tion on their think­ing (which in­cludes the defi­ni­tion I’ve quoted). ↩︎

  3. Dis­claimer: I’m not an economist. I ma­jored in eco­nomics in col­lege and was a re­search as­sis­tant for a few pro­fes­sors, though that was nearly two decades ago. I was ad­mit­tedly fuzzy on some de­tails so I watched a few re­fresher videos. I highly recom­mend Econ­plusDal’s videos (which have mil­lions of views) and have tried to lean on his ex­per­tise where pos­si­ble. For in­stance, when I dis­cuss fac­tors that drive shifts in the la­bor sup­ply curve, the fac­tors I dis­cuss are ones he em­pha­sizes in his video on the topic. I also got very helpful feed­back from an economist on an early draft. ↩︎

  4. This was al­most cer­tainly true in 2015 when EA or­ga­ni­za­tions paid less and tal­ent gaps first started be­ing talked about. ↩︎

  5. Start­ing in the up­per left hand cor­ner, the first ques­tion to ask is whether pay is com­pet­i­tive vs. the for-profit sec­tor. If so, fol­low the “Yes” ar­row to ar­rive at a di­ag­no­sis of a re­cruit­ing prob­lem that sug­gests a need to in­vest more in re­cruit­ing. If that sug­ges­tion worked (“it worked”), the tal­ent con­straint is re­solved. If “it didn’t work”, then the sug­ges­tion is to try in­vest­ing even more in re­cruit­ing. If there’s no money available, we fol­low the “need $” ar­row to ar­rive at a new di­ag­no­sis of a fund­ing prob­lem. ↩︎

  6. 80K’s most re­cent ad­vice on this is­sue is: “In gen­eral, if an or­ga­ni­za­tion thinks that it’s con­strained by a par­tic­u­lar role, we think they should se­ri­ously con­sider rais­ing salaries for that role. How­ever, we don’t think this will fully solve the prob­lem, and it isn’t always the right thing to do.” ↩︎

  7. To be clear, the liter­a­ture on non­profit com­pen­sa­tion in­cludes sig­nifi­cant dis­cus­sion of how the “moral­ity vs. fru­gal­ity trade­off” im­pacts la­bor sup­ply un­der differ­ent con­di­tions. For ex­am­ple, this ar­ti­cle syn­the­sizes em­piri­cal and the­o­ret­i­cal find­ings about the non­profit wage gap: Hal­lock, K. (2000). Com­pen­sa­tion in non­profit or­ga­ni­za­tions [Elec­tronic ver­sion]. Re­search in Per­son­nel and Hu­man Re­sources Man­age­ment, 19, 243-294. http://​​digi­tal­com­mons.ilr.cor­nell.edu/​​hrpubs/​​18/​​ ↩︎

  8. A short­age of ex­pe­rienced tal­ent seems clearly prob­le­matic. One could ar­gue a hy­per-com­pet­i­tive job mar­ket is a pos­i­tive thing, but anec­dotes like this are ob­vi­ously wor­ri­some: “I’ve re­cently grad­u­ated from one of the top ~10 uni­ver­si­ties wor­ld­wide, af­ter in­vest­ing heav­ily in EA through­out my stud­ies. While a stu­dent, EA was the biggest thing in my life… Over the last seven months, I’ve made over 20 un­suc­cess­ful job ap­pli­ca­tions (I keep a spread­sheet). This has in­creased the sever­ity of my de­pres­sion and anx­iety. Over time, I be­gan to shed my iden­tity as an EA, no doubt as a self-defence mechanism.” ↩︎

  9. I’d also sug­gest sur­vey­ing a broader set of or­ga­ni­za­tions. The lat­est sur­vey in­cluded some changes that par­tially ad­dressed con­cerns that have been raised about the po­ten­tial for se­ri­ous se­lec­tion bias, but much more could be done in this di­rec­tion. That said, I rec­og­nize the EA com­mu­nity has some ex­perts on sur­vey de­sign (Re­think Char­ity has done a ter­rific job with the EA Sur­vey for in­stance), and would wel­come (and gen­er­ally defer to) their sug­ges­tions. ↩︎

  10. The 2017 EA Lead­ers Sur­vey asked re­spon­dents “If you tried to spend more money on at­tract­ing tal­ent what might you do?” While most re­spon­dents pro­vided ideas (typ­i­cally rais­ing salaries and/​or in­vest­ing in re­cruit­ing), about ⅓ noted con­cerns that a higher salary for a new hire would re­quire rais­ing salaries for ex­ist­ing staff to avoid re­sent­ment. My sug­gested new ques­tion in­tends to get more ex­plicit and quan­tified in­for­ma­tion about the ex­tent to which this dy­namic is shap­ing EA hiring de­ci­sions and or­ga­ni­za­tional growth tra­jec­to­ries. ↩︎

  11. To make these com­par­i­sons, it’s gen­er­ally helpful to dis­cuss salaries via a cer­tain sized dis­count (pre­mium) to mar­ket, in per­centage terms. For ex­am­ple one might say: “Or­ga­ni­za­tion X pays se­nior de­vel­op­ers 20% more than what other non­prof­its pay, but that’s a 30% dis­count to the open mar­ket so they’re hav­ing trou­ble hiring.” If that mes­sage were ex­pressed in dol­lar terms (rather than per­centages), it’d be less com­pa­rable to analo­gous mes­sages about differ­ent roles, ge­ogra­phies, etc. ↩︎

  12. Ask­ing or­ga­ni­za­tions how much fi­nan­cial value they place on their last hire ap­pears par­tic­u­larly con­fus­ing and prob­le­matic. I’d also sug­gest dis­con­tin­u­ing the “dis­count rate” ap­proach to mea­sur­ing hiring ur­gency since it is yield­ing re­sults that are too noisy to act on. In­stead, I’d ask or­ga­ni­za­tions to di­rectly rate how ur­gent they find hiring on a Lik­ert scale. ↩︎

  13. Here’s a sim­plified ex­am­ple of my in­tu­ition on the right strat­egy for the EA ecosys­tem to adopt. Imag­ine an EA or­ga­ni­za­tion is try­ing to make an im­por­tant hire with skills that are highly de­sired by the for-profit sec­tor. If good can­di­dates could earn $75,000/​year el­se­where in the non­profit sec­tor, and $150,000 in the for-profit sec­tor, ini­tially offer­ing in the neigh­bor­hood of $100,000 could be a good strat­egy. At that salary, the EA char­ity would be ex­tremely at­trac­tive to can­di­dates strongly lean­ing to­ward non­profit work. And while can­di­dates would still need to take a sub­stan­tial $50,000/​year dis­count rel­a­tive to what they could earn in the for-profit sec­tor, that’s a much smaller dis­count than other non­prof­its would re­quire. So offer­ing $100,000 could at­tract a lot of peo­ple who are on the fence about non­profit and for-profit work (per­haps for rea­sons that have noth­ing to do with their mis­sion al­ign­ment or job fit). If we change the ex­am­ple to some­one who could earn $1 mil­lion (or more) in the for-profit sec­tor, my in­tu­ition is much weaker. In prin­ci­ple, I don’t have any ob­jec­tion to non­prof­its pay­ing key staff hun­dreds of thou­sands of dol­lars a year if that makes them more im­pact­ful over­all. ↩︎

  14. Imag­ine a par­allel uni­verse where 80K’s origi­nal in­sights about tal­ent gaps were re­leased with a ti­tle and fram­ing of “Why tal­ent gaps may be more im­por­tant than you think” in­stead of the “Why you should fo­cus more on tal­ent gaps, not fund­ing gaps”. I think the EA com­mu­nity would be more im­pact­ful and on a bet­ter tra­jec­tory in that uni­verse. For what it’s worth, I agree with 80K’s clar­ifi­ca­tion that “it’s pos­si­ble for an or­ga­ni­za­tion to be mainly con­strained by a cer­tain skill, or fund­ing, or both, or nei­ther”. But I don’t think their head­line sum­mary of the is­sue (“skill and fund­ing con­straints are not op­po­sites”) cap­tures the essence of the re­la­tion­ship which I see as “fund­ing gaps are a fre­quent cause of tal­ent gaps, par­tic­u­larly in sec­tors with struc­turally non-com­pet­i­tive salaries”. ↩︎

  15. I’m wary of the­o­ries that as­sume “EA ex­cep­tion­al­ism.” EA or­ga­ni­za­tions aren’t the only ones staffed by peo­ple who be­lieve they are work­ing on the most im­por­tant cause, and are sac­ri­fic­ing earn­ing power to do so. ↩︎