EA Grants applications are now open

I’m an­nounc­ing that the Cen­tre for Effec­tive Altru­ism’s (CEA) EA Grants pro­ject has now been re­opened for pub­lic ap­pli­ca­tions. If you have a pro­ject that you think is worth fund­ing, you can fill out our short ap­pli­ca­tion here. Ap­pli­ca­tions close on Sun­day, Oc­to­ber 14. We ex­pect to run reg­u­lar rounds of pub­lic EA Grant ap­pli­ca­tions like this. We ex­pect the next round af­ter this one to be early next year but we want to re­view les­sons from this round be­fore com­mit­ting to a date.

CEA’s mis­sion is to build the EA com­mu­nity be­cause we be­lieve that a com­mu­nity of ca­pa­ble, ded­i­cated, and al­tru­is­tic peo­ple can make real progress on some of the world’s most sig­nifi­cant prob­lems. As part of this work, we look for ways to help co­or­di­nate the dis­tri­bu­tion of re­sources to the places in the com­mu­nity where they will have the high­est im­pact. In many cases, peo­ple in this com­mu­nity have ideas for promis­ing pro­jects but lack suffi­cient ac­cess to fund­ing to pur­sue their ideas. The goal of EA Grants is to find and sup­port these kinds of pro­jects.

Below I ex­plain the cur­rent sta­tus of EA Grants, the ap­pli­ca­tion pro­cess, our eval­u­a­tion crite­ria and what kinds of pro­jects should ap­ply.

Cur­rent sta­tus of EA Grants

We ran our first pub­lic round of EA Grants in the sum­mer of 2017 and dis­tributed around £500,000. In to­tal this con­sumed ~800 hours of staff time. At its peak, EA Grants re­quired three staff mem­bers work­ing on it on a roughly full time ba­sis. After the ini­tial EA Grants round we con­cluded that we would likely want to run a similar pro­ject again in the fu­ture, prefer­ably at a lower staff time cost than the pre­vi­ous round. You can read a re­view of that round here.

We had planned to open a pub­lic round of Grants in early 2018, but re­al­ized we did not have the staff ca­pac­ity to pro­cess the num­ber of ap­pli­ca­tions we ex­pected. We elected to de­vote staff time to launch­ing EA Com­mu­nity Build­ing Grants and to launch­ing our in­di­vi­d­ual out­reach re­treats (such as our Oper­a­tions Fo­rum) in­stead of de­vot­ing that time to re­open­ing EA Grants for pub­lic ap­pli­ca­tions. When we be­gan our hiring round in the sum­mer we then ad­ver­tised for some­one to take on our EA Grants Eval­u­a­tor role full time. As a stop-gap mea­sure we de­cided to eval­u­ate EA Grants on a refer­ral ba­sis, and we dis­tributed links to the ap­pli­ca­tion form to some mem­bers of the EA com­mu­nity whose judge­ment we trust. This had the benefit of mean­ing we had to re­view fewer ap­pli­ca­tions and so could run the pro­ject with less staff ca­pac­ity.

Through the refer­ral pro­cess, we’ve agreed to fund 22 grants for a to­tal of ~$850,000, us­ing sub­stan­tially less staff time than the 2017 EA Grants round. How­ever, the refer­ral sys­tem has the sig­nifi­cant down­side of mak­ing it less likely that we en­counter pro­jects from peo­ple out­side of our net­works. It also meant that some po­ten­tially promis­ing ap­pli­cants may have failed to de­velop pro­jects that would have been good can­di­dates for EA Grants fund­ing, be­cause they didn’t know that EA Grants fund­ing was still available.

Our ul­ti­mate goal is to keep EA Grants open for any­one to ap­ply on an on­go­ing ba­sis. We’re cur­rently look­ing to hire an EA Grants Eval­u­a­tor to over­see the eval­u­a­tion pro­cess and to give us enough ca­pac­ity to re­li­ably keep EA Grants open for ap­pli­ca­tions. How­ever, even with a full time EA Grants Eval­u­a­tor on staff, when we re­open EA Grants to the pub­lic, we ex­pect to get an ini­tial surge of ap­pli­ca­tions that might prove too much for us to eval­u­ate in a timely man­ner.

To deal with this con­cern and to ac­cel­er­ate the timeline for a pub­lic EA Grants ap­pli­ca­tion, we’ve de­cided to open EA Grants to the pub­lic, but to re­strict the num­ber of ap­pli­ca­tions that move from the first stage of the eval­u­a­tion pro­cess (the short ap­pli­ca­tion) to the sec­ond stage of the pro­cess (the long ap­pli­ca­tion) to 100. This should en­sure that we don’t end up with more ap­pli­ca­tions than we can eval­u­ate in a timely man­ner. We plan to no­tify ap­pli­cants that we sus­pect might have ad­vanced with­out the re­stric­tion and ask them to reap­ply at a later date.

What hap­pens when I ap­ply?

The ap­pli­ca­tion pro­cess for EA Grants now has five ba­sic stages:

1. Short ap­pli­ca­tion
2. Long ap­pli­ca­tion
3. Refer­ence check
4. In­ter­view
5. Fi­nal decision

Ap­pli­ca­tions are open now and will close on Sun­day, Oc­to­ber 14. We ex­pect to be­gin eval­u­at­ing ap­pli­ca­tions be­fore the ap­pli­ca­tion pe­riod ends, so it’s ad­van­ta­geous to sub­mit your ap­pli­ca­tion sooner rather than later.

Between Oc­to­ber 19 and 26, we will no­tify the top 100 ap­pli­cants and ask them to com­plete our long ap­pli­ca­tion. As men­tioned above, we’re re­strict­ing the num­ber of ap­pli­ca­tions that we eval­u­ate fur­ther to en­sure that we have suffi­cient ca­pac­ity to perform the eval­u­a­tions. This means that we ex­pect to re­ject some ap­pli­cants who we think might have been plau­si­ble can­di­dates for fund­ing af­ter fur­ther eval­u­a­tion. After the long ap­pli­ca­tion we will se­lect the top ap­pli­cants for in­ter­views, refer­ence checks and, po­ten­tially, fund­ing. Given the stages re­quired to as­sess grant ap­pli­ca­tions, we ex­pect to make fi­nal de­ci­sions in early De­cem­ber. We’ll keep ap­pli­cants up to date on the progress of their ap­pli­ca­tion by email.

Changes from the pre­vi­ous EA Grants round

The biggest change over the 2017 round is that we broke up the ap­pli­ca­tion form into two stages: the short ap­pli­ca­tion and the long ap­pli­ca­tion. In the 2017 round, we no­ticed that many of the ap­pli­ca­tions could be re­jected af­ter a brief re­view be­cause they weren’t fo­cused on a plau­si­bly high pri­or­ity area or be­cause the ap­pli­cant lacked any signs of al­ign­ment with EA val­ues. We also sus­pect that some plau­si­ble ap­pli­cants failed to ap­ply be­cause the ap­pli­ca­tion form was par­tic­u­larly daunt­ing to com­plete.

The short ap­pli­ca­tion is de­signed to give us enough in­for­ma­tion to de­ter­mine if an ap­pli­ca­tion is plau­si­ble. For ap­pli­ca­tions that aren’t plau­si­ble, the short ap­pli­ca­tion re­duces the time they spend on ap­ply­ing for an EA Grant. For ap­pli­ca­tions that are plau­si­ble, the short ap­pli­ca­tion al­lows us to send a pos­i­tive sig­nal about the ap­pli­ca­tion be­fore ask­ing the ap­pli­cant to spend more time on the ap­pli­ca­tion pro­cess. For the cur­rent ap­pli­ca­tion round, the num­ber of ap­pli­cants that can be moved to the long ap­pli­ca­tion will be re­stricted to 100, but in the fu­ture we plan to ac­cept all ap­pli­ca­tions that seem to be above our fund­ing bar.

We’ve also added a re­quest for three refer­ences to the long ap­pli­ca­tion. We’ll email each refer­ence once the ap­pli­ca­tion is com­plete and ask them to fill out a brief sur­vey about their re­la­tion­ship to the ap­pli­cant. In par­tic­u­larly close fund­ing de­ci­sions we may also call the refer­ences.

Eval­u­a­tion criteria

EA Grants ap­pli­ca­tions are challeng­ing to eval­u­ate. We ac­cept ap­pli­ca­tions in any cause area, and we ac­cept re­quests for many differ­ent types of pro­jects and fund­ing amounts rang­ing from re­quests for a few thou­sand dol­lars to travel to a con­fer­ence to re­quests for a hun­dred thou­sand dol­lars to start a new or­ga­ni­za­tion. In ad­di­tion, be­cause the av­er­age amount of fund­ing re­quested is small, it of­ten doesn’t make sense to spend as much time as we’d like on eval­u­at­ing each re­quest.

The wide range of pos­si­ble pro­jects means that we ex­pect to use some­what differ­ent eval­u­a­tion crite­ria for differ­ent types of pro­jects, and that we ex­pect to eval­u­ate larger fund­ing re­quests much more thor­oughly than smaller fund­ing re­quests. How­ever, we ex­pect the ba­sic eval­u­a­tion frame­work to re­main similar across pro­jects.

Roughly speak­ing, we eval­u­ate pro­jects on the ba­sis of three key ques­tions.

How valuable is this pro­ject?

First, we look at how valuable the pro­ject is likely to be if it goes well. Roughly speak­ing there are three ways a pro­ject can go well: (1) the pro­ject it­self pro­duces some pos­i­tive re­sult; (2) the per­son run­ning the pro­ject gains some valuable skills or ex­pe­rience; (3) some use­ful knowl­edge is cre­ated as a re­sult of the pro­ject.

Ideally pro­jects pro­duce value along mul­ti­ple di­men­sions, but we’re open to fund­ing pro­jects that pro­duce value only on one di­men­sion. For ex­am­ple, we may some­times choose to fund pro­jects where we are un­sure of the ob­ject-level value of the pro­ject, if we think the pro­ject will pro­duce use­ful knowl­edge for the com­mu­nity. While we didn’t fund this par­tic­u­lar pro­ject, an ex­am­ple of a past pro­ject we con­sider worth­while for the in­for­ma­tion it pro­duced was Char­ity Science’s early work on grantwrit­ing for effec­tive char­i­ties, which did not achieve its goal of rais­ing funds but did provide use­ful in­for­ma­tion to other EAs about whether grantwrit­ing on be­half of other or­ga­ni­za­tions was a vi­able pro­ject.

Is this per­son well-suited to this pro­ject?

We also look at whether the ap­pli­cant is a good per­son to run the pro­ject. This breaks down into whether this is a good per­son to run an EA Grants pro­ject in gen­eral, and whether they seem suited to this par­tic­u­lar pro­ject.

The gen­eral char­ac­ter­is­tics we look for in EA Grants ap­pli­ca­tions are un­der­stand­ing of, and com­mit­ment to, the prin­ci­ples of effec­tive al­tru­ism, demon­strated abil­ity and drive, demon­strated abil­ity to run pro­jects au­tonomously, and over­all good judge­ment. For eval­u­at­ing fit be­tween the ap­pli­cant and a par­tic­u­lar pro­ject, we look for ev­i­dence that the ap­pli­cant has the skills and knowl­edge likely to be nec­es­sary to com­plete it.

What hap­pens oth­er­wise?

Fi­nally, we at­tempt to roughly figure out the coun­ter­fac­tu­als to us fund­ing the pro­ject. In par­tic­u­lar, we look at what the per­son would do if the pro­ject isn’t funded and whether they’d be able to ac­cess fund­ing with­out our as­sis­tance. If we think that a per­son’s al­ter­na­tive plans are bet­ter than the pro­ject, then we may dis­cuss that with the ap­pli­cant. If we think the ap­pli­cant can raise money eas­ily from other sources, we may en­courage them to do that in­stead.

Ask­ing what hap­pens oth­er­wise should also pre­dis­pose us in fa­vor of fund­ing peo­ple that are less well known in the EA com­mu­nity or fund­ing pro­jects that fall out­side of es­tab­lished cause ar­eas, as these pro­jects are sig­nifi­cantly less likely to be funded oth­er­wise.

What kinds of pro­jects should ap­ply?

EA Grants has an in­ten­tion­ally broad man­date. We’re open to fund­ing any pro­ject in any cause area. We ex­pect to mostly fund less-es­tab­lished pro­jects with­out easy ac­cess to other fund­ing sources and we gen­er­ally ex­pect to com­mit less than $130,000 per pro­ject, al­though we may make ex­cep­tions in some cases.

One area we’re not cur­rently fund­ing is di­rectly cov­er­ing tu­ition ex­penses for peo­ple who want to go to grad school, al­though we are will­ing to fund re­lated ex­penses, like liv­ing ex­penses to take a high-im­pact in­tern­ship.

The pro­jects that we funded in 2017 are available here. Some ex­am­ple pro­jects that we’ve funded more re­cently in­clude:

- Fund­ing for a pro­ject work­ing on pro­duc­tivity coach­ing for EAs
- Fund­ing for some re­cent grad­u­ates to spend 6 months work­ing out what cause area they should work on by work­ing on some short in­de­pen­dent pro­jects.
- Fund­ing to help a re­searcher tran­si­tion from tra­di­tional poli­ti­cal sci­ence into AI poli­tics and strat­egy.
- Fund­ing to sup­port Char­ity En­trepreneur­ship in cre­at­ing new high-im­pact char­i­ties.
- Fund­ing for re­search on the psy­chol­ogy of effec­tive and in­effec­tive al­tru­ism.

If you have a pro­ject that you think is worth fund­ing, you can fill out our short ap­pli­ca­tion here.