EA Funds—An update from CEA

CEA sees the EA Funds plat­form as a key piece of com­mu­nity in­fras­truc­ture, not only for the con­ve­nience offered to donors who wish to give to one of the four man­aged Funds, but also for its abil­ity to take tax de­ductible dona­tions in the UK and US (the coun­tries where the Cen­tre for Effec­tive Altru­ism is a reg­istered non-profit) and re­grant them to many of the or­ga­ni­za­tions in the EA space.

We are now ded­i­cat­ing the re­sources re­quired to im­prove and build on the early suc­cess of the plat­form (though we rec­og­nize this has not been the case through­out the timeline of the pro­ject). This post will go into de­tail on our plans for fur­ther deep­en­ing the value the EA Funds provide to the EA com­mu­nity, iden­tify mis­takes we have made along the way and provide a gen­eral up­date on our think­ing re­gard­ing the fu­ture of EA Funds.

Ini­tial Context

EA Funds was built dur­ing the early months of 2017, dur­ing CEA’s time as a par­ti­ci­pant in the Y Com­bi­na­tor startup ac­cel­er­a­tor. The plat­form was built as a means to fa­cil­i­tate dona­tions in the EA com­mu­nity. This was built up from (amongst other things) ex­pe­rience with the Giv­ing What We Can Trust, which performed a similar role, but which lacked some of EA Funds’ key fea­tures, and had a smaller sub­set of pos­si­ble grantees.

In par­tic­u­lar, we were mo­ti­vated by the idea that peo­ple of­ten set up their dona­tions on an on­go­ing, re­cur­ring ba­sis (which is good, be­cause it re­moves triv­ial in­con­ve­niences to donat­ing), but don’t of­ten re­view them. We were also in­ter­ested in in­creas­ing spe­cial­iza­tion of dona­tion de­ci­sions, such that a smaller num­ber of more ex­pert peo­ple could de­cide the best ways to spend money, rather than ev­ery­one du­pli­cat­ing the effort of in-depth re­search.

To a large ex­tent, this has been very suc­cess­ful. To date we have re­granted (or are cur­rently in the pro­cess of re­grant­ing) more than $5 mil­lion of EA-al­igned money to around 50 or­ga­ni­za­tions. We’ve been able to ex­tend the plat­form to provide ad­di­tional pay­ment-re­lated ser­vices (for ex­am­ple, the donor lot­tery run at the be­gin­ning of the year), and – in the wake of the crypto boom at the end of 2017 – made it straight­for­ward for donors to donate large cryp­tocur­rency wind­falls tax-effi­ciently.

How­ever, the pro­ject has always ex­isted alongside other CEA pro­jects, and ear­lier in 2018 (and given the plat­form had been op­er­at­ing sta­bly for close to a year), we made the choice to de­pri­ori­tize ac­tive work on EA Funds in favour of de­vot­ing more staff re­sources to other pro­jects (where ac­tive work in­cludes tech­ni­cal work to im­prove the user ex­pe­rience, op­er­a­tions work to e.g. bring on new grantee or­ga­ni­za­tions or to check in on a reg­u­lar ba­sis with Fund man­agers). EA Funds has been op­er­a­tional and has been mak­ing re­grants for the en­tire pe­riod, but it hasn’t been the pri­mary fo­cus of our tech­ni­cal or op­er­a­tions staff.

Our in­ten­tion has always been to come back to EA Funds, and when we made the choice to pri­ori­tize other pro­jects in Q1, we ex­pected to that this would hap­pen later in the year. Fur­ther­more, at the most re­cent CEA team re­treat (mid-June), we re­viewed how EA Funds was op­er­at­ing, and de­cided that we were ready to re-com­mence work on the plat­form in or­der to en­sure that it re­mained a trusted and re­li­able dona­tion op­tion. As such, we made the choice that we would be­gin again in earnest to make some over­due im­prove­ments to the op­er­a­tion of the plat­form, and have been do­ing so since the be­gin­ning of Q3. In other words, we’ve been aware of the same prob­lems that have re­cently been raised in com­mu­nity dis­cus­sion and work­ing on ap­pro­pri­ate solu­tions.

We re­gret not com­mu­ni­cat­ing about this situ­a­tion more openly, and in ret­ro­spect rec­og­nize that wait­ing un­til we have re­solved all the is­sues and had more con­crete, pos­i­tive an­nounce­ments to make has led to a situ­a­tion where some in the com­mu­nity be­gan de­vel­op­ing reser­va­tions about how the plat­form was be­ing run. We plan to be more open about what we are work­ing on with the plat­form go­ing for­ward.

Re­quested improvements

The two main things that com­mu­nity mem­bers have asked for dur­ing the pe­riod EA Funds has been op­er­at­ing have been:

  • Easy visi­bil­ity of cur­rent Fund balances

  • A more con­sis­tent re­grant­ing schedule

The former is dis­cussed in de­tail be­low un­der the head­ing Up­dates to Fund Man­age­ment, the lat­ter un­der Oper­a­tional Up­dates.

Other fea­tures we’re work­ing on at the mo­ment:

  • Bet­ter in­te­gra­tion be­tween pay­roll giv­ing providers and EA Funds (so that soon dona­tions cre­ated through a pay­roll giv­ing provider will be visi­ble in your EA Funds dash­board)

  • An au­to­mated sys­tem to let Fund man­agers and grantees know their cur­rent ac­count bal­ance on a quar­terly ba­sis, to prompt more reg­u­lar regranting

Up­dates to Fund management

We’re cur­rently in the pro­cess of fi­nal­iz­ing some changes to how the Funds are man­aged. The most sig­nifi­cant an­nounce­ment is that Nick Beck­stead will step down as the man­ager of the EA Com­mu­nity and Long-Term Fu­ture Funds.

Nick Beck­stead has pro­vided the fol­low­ing state­ment, ex­plain­ing his rea­sons for step­ping down as a man­ager of these Funds:

At the time I drafted my April up­date, I told CEA that I was in­ter­ested in step­ping down as a man­ager of the EA Com­mu­nity and Long-Term Fu­ture Funds. My rea­sons are as fol­lows.

The origi­nal premise of me as a fund man­ager was that I would fund the grants that I wanted to make but couldn’t fund through Open Phil, and that this wouldn’t re­quire a sub­stan­tial ad­di­tional time in­vest­ment on my part. How­ever, I en­coun­tered fewer grant­mak­ing op­por­tu­ni­ties that I couldn’t fund through Open Phil than I ex­pected, partly be­cause Open Phil’s fund­ing in these ar­eas grew more quickly than I ex­pected. The main thing I’ve wanted to do at Open Phil, but haven’t been able to, is to in­crease the size of my EA and GCR (Global Catas­trophic Risks) grants, which is a large part of the rea­son I made the set of grants that I did. I could in the­ory have spent ad­di­tional time sourc­ing and eval­u­at­ing other op­por­tu­ni­ties, but it has gen­er­ally seemed that ad­di­tional time would be bet­ter spent on grant­mak­ing at Open Phil or other Open Phil work (es­pe­cially given that other fun­ders who have more time to de­vote to the effort and cur­rently are work­ing with smaller bud­gets—such as EA Grants and BERI—were be­com­ing in­volved in the space).

Go­ing for­ward, I be­lieve it will be more im­pact­ful and more satis­fy­ing to our com­mu­nity to have peo­ple man­ag­ing the EA Com­mu­nity and Long-term Fu­ture Funds who can’t make most of their grants through other means, and for whom it would make more sense to de­vote more sub­stan­tial at­ten­tion to the man­age­ment of these funds. I be­lieve this could also be benefi­cial be­cause it would in­crease the di­ver­sity of per­spec­tives of grant­mak­ers and offer oth­ers an op­por­tu­nity to gain ex­pe­rience and es­tab­lish larger track records as grant­mak­ers.

Nick and CEA agreed that he would step down from his role as man­ager of these two Funds af­ter mak­ing his fi­nal set of grants in July 2018, and in late May 2018 a pro­cess for nam­ing his suc­ces­sors was started. Nick has now recom­mended his fi­nal grants, and we are in late stages of the pro­cess to name the new man­agers for these Funds.

Dur­ing this pro­cess we have been fo­cus­ing on find­ing man­agers that are widely re­spected in the com­mu­nity, have some ex­pe­rience with fund­ing rele­vant or­gani­sa­tions at mod­er­ate scale and are able to com­mit the time re­quired to mak­ing grants about twice a year or more. Given some of the con­sid­er­a­tions Nick out­lines above, we’re also plan­ning to ex­per­i­ment with a ‘com­mit­tee’ ap­proach, where sev­eral in­di­vi­d­u­als may be in­volved in pro­vid­ing grant recom­men­da­tions for a par­tic­u­lar Fund. We look for­ward to an­nounc­ing the new man­agers to the com­mu­nity in the com­ing weeks.

It’s worth not­ing that, notwith­stand­ing the change in Fund man­ager, and the fact that we ex­pect this to lead to­wards more reg­u­lar re­grants, we still be­lieve that, as a gen­eral rule, grant man­agers should donate when they see good op­por­tu­ni­ties, not on the ba­sis of a fixed sched­ule, as we think that good grants made later are likely to be bet­ter than rushed grants (at least for some bounded pe­riod of time, af­ter which the effect of dis­count­ing would make the grants equally valuable).

Oper­a­tional Updates

Over the last few months, we have been work­ing on mak­ing im­prove­ments to the plat­form so that the day to day op­er­a­tions run smoothly, trans­par­ently and take a rea­son­able amount of staff time.

The first visi­ble re­sult of this is an up­date to the EA Funds web­site, where each of the four main Funds now has the cur­rent bal­ance (as of the end of the pre­vi­ous cal­en­dar month) dis­played above the list of re­grants, taken di­rectly from our ac­count­ing sys­tem. This will by­pass the need for man­ual up­dates be­ing posted, and al­lows ev­ery­one ac­cess to the lat­est num­bers at all times. If you now view a Fund’s page (see e.g. the An­i­mal Welfare fund’s page ) you’ll see a ‘cur­rent bal­ance’ sec­tion out­lin­ing the amount cur­rently available to the Fund man­ager for re­grant­ing.

We hope to ex­pand on this with more de­tailed re­port­ing func­tion­al­ity in the medium term. Please do note that there will be some vari­abil­ity in the num­bers month to month due to for­eign ex­change fluc­tu­a­tions (bal­ances are held in both USD and GBP) and, at least ini­tially, due to oc­ca­sional ac­count­ing ad­just­ments.

We note that we’ve had the in­ten­tion to get a sys­tem for mak­ing Fund/​Or­ga­ni­za­tion bal­ances visi­ble onto the web­site since early in the year. How­ever, we faced the challenge of pro­vid­ing a defini­tive num­ber for the amount of money in a fund. For a num­ber of fairly mun­dane ac­count­ing rea­sons (e.g. the de­lay be­tween a donor re­port­ing a dona­tion and us re­ceiv­ing the money via bank trans­fer or cheque, donors giv­ing out­side the EA Funds sys­tem, donors chang­ing their dona­tion al­lo­ca­tion, com­pli­ca­tions with ac­count­ing for the ap­pre­ci­ated value of cryp­tocur­rency post-liqui­da­tion and de­ci­sions about whether to use num­bers de­rived on a cash or ac­cru­als ba­sis), it has been difficult for us to point to a sin­gle, canon­i­cal num­ber that ac­cu­rately rep­re­sents the bal­ance of a given Fund at a given point in time.

We’ve spent the bet­ter part of this quar­ter re­fin­ing our pro­cesses so that we now feel con­fi­dent that the num­ber that we’re pul­ling from our ac­count­ing soft­ware is a good rep­re­sen­ta­tion of the ‘cur­rent bal­ance’ of the fund.

A note on differ­ences be­tween March bal­ances and cur­rent balances

The bal­ances as they ap­pear on each Fund’s page are in some cases lower than what the last pub­lic up­date in March (less any grants) would sug­gest. This is due to the March num­bers in­clud­ing ac­counts payable; money not yet held as cash but ex­pected to come in. The largest differ­ence is in the An­i­mal Welfare Fund’s bal­ance, where over $500,000 of such trans­ac­tions was re­ported as part of the bal­ance in March.

It ap­pears that a num­ber of large dona­tions which were re­ported dur­ing the De­cem­ber ‘giv­ing sea­son’ turned out to be peo­ple test­ing/​ex­per­i­ment­ing with the EA Funds plat­form; there­fore, the March bal­ances in­cluded amounts which were never meant as gen­uine dona­tions. While it’s stan­dard ac­count­ing prac­tice to re­port ac­counts re­ceiv­able as an as­set (and then perform reg­u­lar write-offs for money that doesn’t come in), we rec­og­nize this has caused con­fu­sion. As such, the ac­count bal­ances dis­played on the site now ex­clude ac­counts re­ceiv­able to pre­vent this con­fu­sion go­ing for­ward.

We will also con­tinue our work on the ac­counts in col­lab­o­ra­tion with our ex­ter­nal ac­count­ing teams in the UK and US to re­duce the need for core CEA staff to spend too much time on this, and we are de­vel­op­ing new ap­proaches to our re­grant­ing op­er­a­tions, which are cur­rently a highly man­ual pro­cess. This means that we plan to pro­cess all re­grants at the same time once per quar­ter.

We think the EA Funds is uniquely po­si­tioned to provide value to the EA com­mu­nity by sav­ing the du­pli­ca­tion of similar op­er­a­tions effort at other or­gani­sa­tions, and we’re ex­cited to keep im­prov­ing the plat­form for the benefit of the com­mu­nity.

Mes­sag­ing re­gard­ing effec­tive­ness of dona­tions to the man­aged Funds

We ini­tially be­lieved that the four man­aged Funds would sig­nifi­cantly im­prove the effec­tive­ness of the av­er­age EA dona­tion. This was in part based on the as­sump­tion that our Fund Man­agers, drawn from GiveWell and the Open Philan­thropy Pro­ject, would be able to eas­ily find fund­ing op­por­tu­ni­ties that they be­lieved to be on par with their ‘in­sti­tu­tional’ grants, but that their re­spec­tive in­sti­tu­tions would be more ret­i­cent to fund. We have since up­dated our think­ing on this, be­cause this has turned out not to be the case in all situ­a­tions and also be­cause we are now more wor­ried about the effect that hav­ing Fund Man­agers ex­clu­sively from the Open Philan­thropy Pro­ject and GiveWell has on the di­ver­sity of view­points rep­re­sented in fund­ing de­ci­sions within the com­mu­nity.

We are chang­ing the mes­sag­ing on the EA Funds site to bet­ter re­flect this up­dated think­ing. We con­tinue to see the man­aged Funds as an ex­cel­lent dona­tion choice for donors who wish to save time and just want to donate whilst know­ing their dona­tion effec­tive­ness will meet a very high bar. How­ever, for donors more deeply en­gaged in the com­mu­nity it is com­pletely plau­si­ble they are able to out­perform the effec­tive­ness of the man­aged Funds, es­pe­cially in situ­a­tions where they are in a unique po­si­tion to fund an early stage pro­ject they are aware of due to their net­work, or other similar situ­a­tions.

Fur­ther di­ver­sifi­ca­tion of fund­ing view­points: an ex­tra tier of Funds

We have also been plan­ning to fur­ther in­crease the di­ver­sity of views rep­re­sented in the man­aged Funds model by cre­at­ing a new ‘tier’ of Funds. Plans for this are still in early stages, and we’re not plan­ning to in­crease the op­er­a­tional com­plex­ity of the plat­form un­til we have han­dled our ad­minis­tra­tive back­log dis­cussed above. How­ever we hope to be able to an­nounce more de­tailed plans for this in the com­ing months.

The ba­sic idea would be to al­low less-tested grant mak­ers, with par­tic­u­larly rele­vant per­sonal net­works or po­ten­tially less-main­stream views to man­age a ‘Ju­nior Fund’ (work­ing ti­tle), where these would not be pro­moted to the same ex­tent the main four Funds are, but would nev­er­the­less be eas­ily ac­cessible to in­ter­ested donors. We would then ‘let the mar­ket de­cide’ to a cer­tain ex­tent on which of these suc­ceed, set­ting a min­i­mum dona­tion thresh­old to be reached within the first 12 months for these Funds to con­tinue to be worth­while to op­er­ate. It seems plau­si­ble that we could dis­cover highly com­pe­tent new grant-mak­ers this way.

We will most likely run a pub­lic con­test for these po­si­tions.


We would like to seek your feed­back on the plans and up­dates laid out in this post, and I would like to en­courage any­one who would like to dis­cuss any of these mat­ters in more de­tail to reach out to me on funds at effec­tivealtru­ism dot org.