To Grow a Healthy Movement, Pick the Low-Hanging Fruit

This post is co-au­thored by Nick Fitz and Ari Ka­gan and is cross-posted from The Life You Can Save’s blog.

We’ve all met peo­ple who will never be open to the ideas of effec­tive al­tru­ism (EA) no mat­ter the fram­ing: your neigh­bor who thinks that do­ing good is purely a mat­ter of per­sonal prefer­ence or your un­cle who ar­gues it’s wrong to com­pare char­i­ties be­cause they’re all do­ing good. De­spite hours of dis­cus­sion, no real progress is made. We’ve all also had the op­po­site ex­pe­rience: your friend who’s on board a minute into the con­ver­sa­tion; the one who was ex­cited to learn that oth­ers thought the same way that she already did. You were likely such a per­son your­self – you merely needed to learn of the move­ment to be­come a sup­porter.

Much of the de­bate about how to build the EA move­ment is fo­cused on how to frame is­sues to con­vince peo­ple of the mer­its of the gen­eral EA wor­ld­view. Should we talk of giv­ing as an op­por­tu­nity or an obli­ga­tion? Should we em­pha­size global health and poverty over ar­tifi­cial in­tel­li­gence and an­i­mal ethics? Such dis­cus­sions are vi­tal, but craft­ing our mes­sage can only get us so far. Many po­ten­tially highly-en­gaged effec­tive al­tru­ists (EAs) haven’t even heard of EA yet. If we knew who these folks were, we could grow the move­ment far more quickly and sus­tain­ably. It’s far more effec­tive to pri­ori­tize iden­ti­fy­ing peo­ple who are already highly pre­dis­posed to­wards the tenets of effec­tive al­tru­ism than it is to strug­gle with those who are a much harder “sell.”

In the ”Aware­ness/​In­cli­na­tion” model de­vel­oped by the Cen­tre for Effec­tive Altru­ism (CEA), Cot­ton-Bar­ratt pro­poses that move­ment-build­ing is made up of two el­e­ments: mak­ing peo­ple more aware of EA ideas and mak­ing peo­ple more in­clined to­wards EA ideas. CEA notes that it’s much harder to in­crease in­cli­na­tion than aware­ness: “Our cur­rent way of re­solv­ing this is to fo­cus our efforts on in­creas­ing the aware­ness of those who already have rel­a­tively high in­cli­na­tion.” This isn’t to say that we should be closed to peo­ple rad­i­cally chang­ing their be­liefs (quite the op­po­site), but let’s pri­ori­tize the low-hang­ing fruit.

To do so, we need to find peo­ple who are highly in­clined to­wards EA, yet cur­rently un­aware of EA ideas. In­deed, a forth­com­ing agenda from The Life You Can Save ex­plic­itly aims to tar­get “de­mo­graph­ics that are likely to be re­cep­tive to effec­tive giv­ing con­cepts rel­a­tive to the gen­eral pop­u­la­tion.” We’ve all come to as­sume it’s the white, athe­ist, male, de­vel­oper in the bay who listens to Sam Har­ris and reads Less Wrong. Per­haps for good rea­son: the 2017 EA sur­vey found that the cur­rent move­ment is 89% white, 81% athe­ist, 81% ages 20-35, and 70% male. But are these num­bers rep­re­sen­ta­tive of those re­cep­tive to effec­tive al­tru­ism? Or do they sim­ply re­veal how EA has his­tor­i­cally spread through ho­mophilic so­cial net­works, soft­ware en­g­ineer­ing teams, and elite uni­ver­si­ties.

The study

In a re­cent study, we’ve ex­plored these ques­tions to iden­tify who may already be re­cep­tive to EA. We ran an on­line study with 530 Amer­i­cans to de­ter­mine what pre­dicts sup­port for EA. After par­ti­ci­pants read a gen­eral de­scrip­tion of EA, they com­pleted mea­sures of their sup­port for EA (e.g., at­ti­tudes and giv­ing be­hav­iors). Fi­nally, par­ti­ci­pants an­swered a col­lec­tion of ques­tions mea­sur­ing their be­liefs, val­ues, be­hav­iors, de­mo­graphic traits, and more.

The re­sults sug­gest that the EA move­ment may be miss­ing a much wider pop­u­la­tion of highly-en­gaged sup­port­ers. For ex­am­ple, not only were women more al­tru­is­tic in gen­eral (a widely repli­cated find­ing), but they were also more sup­port­ive of EA speci­fi­cally (even when con­trol­ling for gen­eros­ity). And whites, athe­ists, and young peo­ple were no more likely to sup­port EA than av­er­age. If any­thing, be­ing black or Chris­tian in­di­cated a higher like­li­hood of sup­port­ing EA. More­over, the typ­i­cal stereo­type of the “EA per­son­al­ity” may be some­what mis­guided. Many peo­ple—both within and out­side the com­mu­nity—view EAs as cold, calcu­lat­ing types who use ra­tio­nal­ity to over­ride their emo­tions—the sort of peo­ple who can eas­ily ig­nore the beg­gar on the street. Yet the data sug­gest that the more em­pa­thetic some­one is (in both cog­ni­tion and af­fect), the more likely they are to sup­port EA. Im­por­tantly, an­other key pre­dic­tor was the psy­cholog­i­cal trait of ‘max­i­miz­ing ten­dency,’ a de­sire to op­ti­mize for the best op­tion when mak­ing de­ci­sions (rather than set­tle for some­thing good enough). That is, it’s not enough to just care about oth­ers, or to only have a ten­dency to op­ti­mize. When a per­son scores high on both em­pa­thy and max­i­miz­ing, they’re much more likely to en­dorse EA. We set out to in­crease the re­turn on in­vest­ment of move­ment-build­ing work; in do­ing so, we dis­cov­ered an un­der­ex­plored, much more di­verse group in­ter­ested in EA, and found that the EA per­son­al­ity may be some­what mi­s­un­der­stood. For more on these re­sults, see this brief sum­mary.

Is move­ment build­ing the right ap­proach?

Most agree that the move­ment needs more highly com­mit­ted sup­port­ers, but the ques­tions is how to ac­quire them. There are con­cerns that move­ment build­ing may lead to “move­ment dilu­tion,” the worry “that an overly sim­plis­tic ver­sion of EA, or a less al­igned group of in­di­vi­d­u­als, could come to dom­i­nate the com­mu­nity, limit­ing our abil­ity to fo­cus on the most im­por­tant prob­lems.” To avoid this prob­lem, some sug­gest that we should in­stead fo­cus on in­creas­ing the com­mit­ment of ex­ist­ing EAs. While nei­ther fo­cus pre­cludes the other, tra­di­tional move­ment build­ing ap­proaches may have thus far pri­mar­ily at­tracted low-al­ign­ment sup­port­ers, mak­ing it less valuable. Spread­ing ideas through the mass me­dia is a low fidelity way to com­mu­ni­cate, and fram­ing EA to bet­ter con­vince peo­ple may only per­suade those who aren’t closely al­igned. The ap­proach we ad­vo­cate for here sidesteps these wor­ries: ev­i­dence-based move­ment-build­ing can help us iden­tify highly-al­igned sup­port­ers. The min­i­mal effort re­quired to in­crease the aware­ness of some­one who is already highly-in­clined may im­me­di­ately re­sult in a high level of con­tri­bu­tion to the move­ment. The full dis­cus­sion is out­side the scope of this post, but this trade-off mer­its more eval­u­a­tion.

Where do we go from here?

This work is just the tip of the ice­berg, and there’s much more to sort out. Not enough work has been done to craft ev­i­dence-based mod­els of sus­tain­able move­ment-build­ing. Some or­ga­ni­za­tions are start­ing to build these tools: Stu­dents for High-Im­pact Char­ity (SHIC) has started to “col­lect a va­ri­ety of met­rics to find stu­dents most in­clined to en­gage with effec­tive char­ity in the long-term” and the Giv­ing Game Pro­ject is also col­lect­ing sur­vey data to­ward a similar end. Once we have enough data on key traits pre­dict­ing open­ness to the ideas of EA, we should work to find groups of peo­ple that are high in these traits and fo­cus our efforts where there will be greater re­turn. We’d love for this to snow­ball in the com­mu­nity as the EA move­ment in­cor­po­rates data-driven de­ci­sion-mak­ing more gen­er­ally.

If the EA move­ment bet­ter un­der­stands who may already be in­ter­ested in what it has to say, then it will be able to grow sus­tain­ably at a much faster rate. This may seem ob­vi­ous, but it is not the cur­rent ap­proach. This could change not only who we try to en­gage with effec­tive al­tru­ism, but how we hire peo­ple, how we run fundrais­ing cam­paigns, and more. It’s time to un­der­stand who tends to sup­port effec­tive al­tru­ism, and why. By em­brac­ing ev­i­dence-based ap­proaches to build­ing the ev­i­dence-based al­tru­ism move­ment, we can do even more good with our re­sources.

If you’re in­ter­ested in learn­ing more about the study, we wel­come emails at ar­ink­a­ and