GiveWell money moved forecasts and implications

Note: See UPDATES 1, 2, 3, 4, and 5 at the bot­tom of the post.

This post as­sumes some fa­mil­iar­ity with the work of char­ity eval­u­a­tor GiveWell, but here’s a quick back­grounder for those not fa­mil­iar: GiveWell lists a small set of top char­i­ties to donate to, and pub­lishes de­tailed re­views of each. The list for 2015 com­prises Against Malaria Foun­da­tion (malaria bed­net dis­tri­bu­tion), GiveDirectly (un­con­di­tional cash trans­fers), De­worm the World Ini­ti­a­tive (de­worm­ing), and Schis­to­so­mi­a­sis Con­trol Ini­ti­a­tive (de­worm­ing). GiveWell’s recom­men­da­tions have ac­quired the sta­tus of a gold stan­dard for con­ven­tional, unattached donors, and GiveWell has a par­tic­u­larly strong rep­u­ta­tion among many of the peo­ple who read this site and iden­tify as effec­tive al­tru­ists. GiveWell also works closely with Good Ven­tures, the pri­vate foun­da­tion of Face­book co-founder Dustin Moskovitz and his wife Cari Tuna. Good Ven­tures has made sig­nifi­cant dona­tions to GiveWell’s top char­i­ties, mostly at GiveWell’s sug­ges­tion, with the amount reach­ing $44.4 mil­lion (plus $1 mil­lion to stand­out char­i­ties, plus a sep­a­rate three-year $25 mil­lion grant to GiveDirectly) in 2015.

In a re­cent post on the Effec­tive Altru­ists Face­book group cri­tiquing what I saw as claims of take-off ex­po­nen­tial growth of the EA move­ment, I noted some es­ti­mates of what I be­lieve GiveWell’s money moved will be for 2015 and the years to come. In this post, I want to de­velop and re­fine those es­ti­mates more.

TL;DR

[Added at re­quest of com­menters, even though I’m skep­ti­cal that this TL;DR can ac­cu­rately cap­ture the en­tirety of the post].

For sim­plic­ity, this TL;DR does not de­scribe the in­ter­vals of un­cer­tainty. For those you’ll have to read the whole post (or the ap­pro­pri­ate sec­tion thereof).

  • Fore­cast­ing the amount of money moved by GiveWell helps donors un­der­stand their marginal im­pact. It also helps re­cip­i­ent or­ga­ni­za­tions and fundraisers.

  • GiveWell and its recom­mended char­i­ties are cur­rently a very small frac­tion of the sec­tors within which they op­er­ate. There are no hard ceilings be­ing hit just yet.

  • I es­ti­mate GiveWell’s non-Good Ven­tures money moved to grow by a lit­tle over 40% from 2014 to 2015, though there is con­sid­er­able un­cer­tainty. I be­lieve that Elie’s es­ti­mate of well over 20 mil­lion dol­lars in money moved is a tad over­con­fi­dent.

  • I be­lieve that GiveWell’s growth in money moved this year will come partly from an in­crease in the num­ber of donors (about 20%) and partly from an in­crease in the amount of money donated per donor (about 20%). In fu­ture years, growth will be con­cen­trated more in terms of money moved per donor rather than num­ber of donors, though both will grow mod­estly.

  • I see three key de­ter­mi­nants of GiveWell’s money moved: dis­cov­ery/​brand aware­ness, con­ver­sion, and re­ten­tion/​up­sell. The biggest con­straint on GiveWell’s money moved is poor con­ver­sion. The biggest source of near-term growth is re­ten­tion/​up­sell. While dis­cov­ery/​brand aware­ness is low, in­creas­ing it sig­nifi­cantly won’t help much since con­ver­sion is a big­ger bot­tle­neck and will be an even big­ger bot­tle­neck if brand aware­ness were sud­denly in­creased.

  • GiveWell’s choice of top char­i­ties it­self af­fects its money moved, and the 2015 list of top char­i­ties is most con­ducive to mov­ing more money.

  • The long-run solu­tion to the con­ver­sion prob­lem is ex­pand­ing the range of recom­mended char­i­ties, but this is un­likely to hap­pen in a mean­ingful way for at least the next two years.

Why is this im­por­tant?

GiveWell was an early pro­po­nent of the idea of room for more fund­ing, and of the closely re­lated con­cept of marginal im­pact, i.e., what effect does your marginal con­tri­bu­tion have? In or­der to best es­ti­mate an or­ga­ni­za­tion’s room for more fund­ing, GiveWell tries to es­ti­mate how much the or­ga­ni­za­tion will raise from non-GiveWell donors, and also takes into ac­count the amount it is recom­mend­ing that Good Ven­tures donate. How­ever, GiveWell doesn’t provide offi­cial guidance on how much money it ex­pects GiveWell-in­fluenced donors other than Good Ven­tures to give. I be­lieve that car­ry­ing out this es­ti­ma­tion is im­por­tant for a few rea­sons:

  • It tells donors the rele­vant mar­gin at which their dona­tion is ac­tu­ally hap­pen­ing: One of the biggest de­ter­mi­nants of marginal im­pact is how much money would be moved in the ab­sence of your dona­tion. [Math­e­mat­i­cally, if we think of marginal im­pact in terms of the deriva­tive of the im­pact func­tion, then know­ing how much money would be moved oth­er­wise tells us at what point we need to take the deriva­tive.]

  • It helps re­cip­i­ent or­ga­ni­za­tions plan bet­ter for the fu­ture: Know­ing how much money they ex­pect to raise can help the or­ga­ni­za­tions bud­get ap­pro­pri­ately for staff and pro­grams. Over­es­ti­mates could lead to tak­ing up too many pro­grams that they then can­not fol­low through, whereas un­der­es­ti­mates can lead to sig­nifi­cant op­por­tu­ni­ties be­ing un­der­filled.

  • It helps fundrais­ing or­ga­ni­za­tions and in­di­vi­d­u­als un­der­stand the mar­gins at which they are operating

It could still be the case that while ac­cu­rate an­swers are helpful, wild spec­u­la­tion isn’t. But get­ting to ac­cu­rate an­swers it­self re­quires prac­tice at es­ti­ma­tion and fore­cast­ing, even if ini­tial fore­casts are of low qual­ity. This post is an at­tempt at ini­tial fore­casts, along with a walk­through of the rea­son­ing used to gen­er­ate the fore­casts. The fore­casts might well be of low qual­ity, but I hope peo­ple find the ex­er­cise illu­mi­na­tive.

In ad­di­tion to the value of the fi­nal fore­casts, the rea­son­ing used in ar­riv­ing at the fore­casts helps shed light on other facets im­por­tant to donors, re­cip­i­ent or­ga­ni­za­tions, char­ity eval­u­a­tors, and fundrais­ing or­ga­ni­za­tions.

The amounts of money moved are now suffi­ciently sig­nifi­cant that this sort of anal­y­sis is worthwhile

As GiveWell’s donor base has in­creased in size and its an­nual money moved has hit eight digits, es­ti­ma­tion makes more and more sense. A few rea­sons:

  • With more donors and larger amounts, more ac­cu­rate fore­cast­ing be­comes more pos­si­ble (with the caveat that a billion­aire here or there could tip the scales). Ex­plic­itly, with a larger num­ber of donors, the law of large num­bers starts kick­ing in. More­over, if donors are donat­ing larger amounts, the pro­por­tional vari­a­tion in the amounts they donate is likely to be less (be­cause the differ­ence be­tween donat­ing $10,000 and $20,000 for a donor is more sig­nifi­cant than the differ­ence be­tween donat­ing $10 and $20).

  • The re­cip­i­ent or­ga­ni­za­tions have now demon­strated room for more fund­ing ac­cord­ing to GiveWell, and are plan­ning their ex­pan­sion op­er­a­tions speci­fi­cally with the ex­pec­ta­tion of rais­ing money from GiveWell. In par­tic­u­lar, Against Malaria Foun­da­tion and GiveDirectly seem to be ex­pand­ing rapidly with GiveWell play­ing a key role in their growth. While a lot of GiveWell’s con­tri­bu­tion comes through Good Ven­tures grants, in­di­vi­d­ual donors play a non­triv­ial part for both or­ga­ni­za­tions.

  • GiveWell it­self holds con­sid­er­able promise as a pi­o­neer of the idea of effec­tive al­tru­ism (broadly con­strued, not nec­es­sar­ily to be equated with the epony­mous so­cial move­ment or this web­site). Un­der­stand­ing the growth of its money moved can shed in­sight into just how far the idea of effec­tive al­tru­ism can go in the world of philan­thropy.

An over­all sense of magnitude

To give some sense of the num­bers:

  • Money donated to char­ity in the United States is about 350 billion dol­lars a year, which is a lit­tle over 2% of US GDP, which in turn is around 15-20% of Gross World Product.

  • GiveWell’s non-Good Ven­tures money moved to top char­ity was $12.7 mil­lion-$13.0 mil­lion in 2014, which is around 0.004% of to­tal money moved to char­ity in the United States. GiveWell doesn’t offi­cially pub­lish dis­tri­bu­tion of donors by coun­try (I did write to them and they said they cur­rently don’t have that data) but my guess is that over half of its money moved comes from US donors, and pos­si­bly a lot more. So, ex­clud­ing Good Ven­tures, GiveWell is re­spon­si­ble for mov­ing some­where be­tween 0.002% and 0.004% of money moved to char­ity in the United States. Or viewed an­other way, to­tal money moved to GiveWell-recom­mended char­i­ties in 2014, ex­clud­ing money from Good Ven­tures, rep­re­sented 0.00001% of gross world product.

  • Based on var­i­ous sur­veys of EA or­ga­ni­za­tions and data pub­lished by EA or­ga­ni­za­tions, my guess is that some­where be­tween 10% and 30% of money moved by GiveWell was through peo­ple who got to GiveWell through the EA move­ment. I won’t elab­o­rate on this point much here be­cause it’s not the fo­cus of my es­ti­ma­tion ex­er­cise, but you can check out the 2014 EA Sur­vey re­sults for some ex­am­ple data.

What about GiveWell-recom­mended char­i­ties? Here are some rough es­ti­mates I got with a few min­utes of Googling; num­bers may be far from ac­cu­rate.

  • Against Malaria Foun­da­tion works in the do­main of bed­net dis­tri­bu­tion. Prior to 2015, it had raised less than 20 mil­lion dol­lars in to­tal (a large part of it from Good Ven­tures and oth­ers in­fluenced by GiveWell’s recom­men­da­tions, plus some in­fluenced by Giv­ing What We Can recom­men­da­tions). With the new in­flux of Good Ven­tures and GiveWell-in­fluenced money, its an­nual rev­enue could be in the tens of mil­lions of dol­lars. For com­par­i­son, the Global Fund to Fight AIDS, Tu­ber­cu­lo­sis, and Malaria, a Gates Foun­da­tion-backed and in­ter­na­tion­ally sup­ported grant­maker, spends be­tween 2 and 5 billion a year, and de­votes about 20% of its ex­penses to fight­ing malaria (ac­cord­ing to GiveWell’s page on them). The es­ti­mate would there­fore be that it spends be­tween 400 mil­lion and 1 billion dol­lars on anti-malaria pro­grams. Thus, prior to 2015, AMF ac­counted for less than 1% of global philan­thropic ex­pen­di­ture against malaria, whereas it could now go up to any­where be­tween 2% and 10% (de­pend­ing on how the other num­bers shake out).

  • GiveDirectly is the only non-gov­ern­men­tal or­ga­ni­za­tion de­voted to large-scale un­con­di­tional cash trans­fers, and it’s not im­me­di­ately clear what to com­pare it against. As a frac­tion of the econ­omy of the coun­tries it op­er­ates in, it’s small but not as in­signifi­cant as it is for the send­ing economies. For in­stance, its an­nual ex­penses in Kenya are un­der 0.1% of Kenya’s nom­i­nal GDP, but more than 0.01%.

  • I wasn’t able to get good es­ti­mates of the to­tal money spent on de­worm­ing, but since SCI and DtWI (and in par­tic­u­lar DtWI) work with gov­ern­ments, the money lev­er­aged by their efforts is sig­nifi­cantly more than their own money spent.

Th­ese broad es­ti­mates show that in terms of money moved, frac­tion of the philan­thropic sec­tor, frac­tion of the wealth in donor groups, and frac­tion of the re­sources available to re­cip­i­ent groups, GiveWell and its top char­i­ties are still quite small. This sug­gests that there is con­sid­er­able up­ward room to grow with­out hit­ting hard limits. Bar­ri­ers to growth can of course arise in a va­ri­ety of ways, but even so, there is no strong ev­i­dence against the propo­si­tion that, over the next few decades, GiveWell’s an­nual non-Good Ven­tures money moved can grow by 1-2 or­ders of mag­ni­tude, while still re­main­ing a small frac­tion of the quan­tities in­volved. For in­stance, GiveWell cited the Money for Good study (funded to help re­think the Non­profit Mar­ket­place Ini­ti­a­tive of the Hewlett Foun­da­tion) and noted that, based on in­ter­pre­ta­tion, some­where be­tween $550 mil­lion and $4 billion in an­nual giv­ing in the United States could be in­fluenced by GiveWell-like re­sources. This would be some­where be­tween 0.2% and a lit­tle over 1% of an­nual giv­ing in the United States, and rep­re­sents a 40-200X in­crease in the non-Good Ven­tures money moved by GiveWell.

Money moved estimates

A few rele­vant pieces of in­for­ma­tion:

My es­ti­mates for the non-Good Ven­tures money moved by GiveWell have changed slightly since my Effec­tive Altru­ists Face­book group post; here are the re­vised es­ti­mates.

  • Es­ti­mate for 2015: Me­dian es­ti­mate of $18.7 mil­lion, 50% prob­a­bil­ity es­ti­mate of $16.2 mil­lion to $24.5 mil­lion.

  • Es­ti­mate for 2016: Me­dian es­ti­mate of $25.3 mil­lion, 50% prob­a­bil­ity es­ti­mate of $17.5 mil­lion to $37.5 mil­lion.

  • Es­ti­mate for 2017: Me­dian es­ti­mate of $30.1 mil­lion, 50% prob­a­bil­ity es­ti­mate of $19.5 mil­lion to $58 mil­lion.

In con­trast, ac­cord­ing to an Oc­to­ber 2015 board meet­ing, Elie Hassen­feld of GiveWell was con­fi­dent that it would move over 20 mil­lion dol­lars in 2015. I be­lieve this con­fi­dence is mis­placed, for rea­sons I dis­cuss be­low (with that said, the 20 mil­lion+ sce­nario is about 40% likely, so it’s not way off from my es­ti­mate). But it is also pos­si­ble that they have in­sider in­for­ma­tion that I lack ac­cess to. That is part of the rea­son for my up­ward re­vi­sion of my es­ti­mates since the Face­book post.

For the 2015 money moved es­ti­mate, here are some more de­tails be­fore I delve into my rea­sons.

Here is the dis­tri­bu­tion by per­centile:

  • 1st per­centile: $8.8 million

  • 2.5th per­centile: $11.6 million

  • 10th per­centile: $14.7 million

  • 25th per­centile: $16.2 million

  • 50th per­centile: $18.7 million

  • 75th per­centile: $24.5 million

  • 90th per­centile: $28.6 million

  • 97.5th per­centile: $33.1 million

  • 99th per­centile: $45 million

Here is the dis­tri­bu­tion for top char­i­ties (50% prob­a­bil­ity for each):

  • Against Malaria Foun­da­tion: $8 mil­lion - $17 million

  • GiveDirectly: $1 mil­lion - $4 million

  • Schis­to­so­mi­a­sis Con­trol Ini­ti­a­tive: $1 mil­lion - $4 million

  • De­worm the World Ini­ti­a­tive: $250,000 - $4 million

The low lower-end es­ti­mate for De­worm the World Ini­ti­a­tive arises from the fact that it raised less money last year, and the in­creased en­dorse­ment of the or­ga­ni­za­tion hap­pened at the same time that GiveWell dou­bled down on AMF as a top pick. More­over, Good Ven­tures has agreed to fill its en­tire Level 1 fund­ing gap, mak­ing it less at­trac­tive for in­di­vi­d­ual donors who take GiveWell’s anal­y­sis se­ri­ous.

Here’s the size dis­tri­bu­tion of donors for 2012, 2013, and 2014 (from the same GiveWell blog post).

GiveWell money moved by size bucket

We see that 2013 was a sig­nifi­cant in­crease year for num­ber of donors, whereas 2014 was an year when the num­ber of donors didn’t in­crease much (there was a lot of churn) but the av­er­age money moved per donor in­creased sig­nifi­cantly. I be­lieve that the situ­a­tion for 2015 will look pos­i­tive on both fronts: an in­crease in the num­ber of donors, and an in­crease in the money moved per donor. My me­dian es­ti­mates for the growth on both fronts is around 20%. Ex­plic­itly, my 50% in­ter­val for the num­ber of donors in 2015 is 9,500-12,500. My 50% in­ter­val es­ti­mate for the money moved per donor is $1,500-2,000.

Ex­pla­na­tion of the lower and up­per end estimates

Ex­pla­na­tion of the lower end: Since GiveWell has already moved $5.56 mil­lion in money in the first two quar­ters of 2015, and Elie was rea­son­ably con­fi­dent of the 20 mil­lion figure for the whole year, I ex­pect that dona­tions did not com­pletely stop in Q3 of 2015. There­fore, I ex­pect that just the first three quar­ters alone would have ac­counted for at least $6.5 mil­lion in dona­tions. That leaves about $2.3 mil­lion in po­ten­tial dona­tions for the end of year, which is fairly easy to achieve with ex­ist­ing pledges. There­fore, I ex­pect a prob­a­bil­ity of about 1% of miss­ing the $8.8 mil­lion mark. That 1% cov­ers ma­jor scan­dals at GiveWell and its recom­mended char­i­ties, or other un­fore­seen prob­lems.

Ex­pla­na­tion of the up­per end: The up­per end of es­ti­mates is mainly gov­erned by fac­tors such as wealthy donors, highly suc­cess­ful fundrais­ing groups or celebrity en­dorse­ments or me­dia pub­lic­ity, and other un­fore­seen fac­tors.

With the lower and up­per end taken care of, I’ll fo­cus on ex­plain­ing the mid­dle. Speci­fi­cally, I want to ex­plain why my me­dian es­ti­mate is lower than the “well over 20 mil­lion” that Elie and Holden have put out.

A quick the­o­ret­i­cal model for how GiveWell moves money

One thing seems rea­son­ably clear to me from many differ­ent sources of anec­do­tal ev­i­dence:

Here’s my de­scrip­tion of the most typ­i­cal path­way for a GiveWell donor:

  • The per­son dis­cov­ers GiveWell through some method such as search, friend refer­ral, so­cial me­dia, or a web­site or per­son they fol­low on­line.

  • The per­son checks out GiveWell’s site, and hears pos­i­tive things about GiveWell.

  • When Giv­ing Sea­son (con­cen­trated around De­cem­ber) ar­rives, the per­son checks out GiveWell’s recom­men­da­tions, and in some cases makes a dona­tion to a GiveWell-recom­mended char­ity on GiveWell’s recom­men­da­tion.

Based on this model, a few key de­ter­mi­nants of GiveWell’s money moved stand out:

  • Dis­cov­ery and brand aware­ness: How many new peo­ple are dis­cov­er­ing GiveWell? What is the to­tal num­ber of peo­ple who know about GiveWell and what it stands for? I did a Face­book post on a Sur­veyMon­key Au­di­ence sur­vey of brand aware­ness of var­i­ous philan­thropic or­ga­ni­za­tions. GiveWell did de­cently okay but way be­low the Gates Foun­da­tion, and also be­low the rel­a­tively re­cent Chan Zucker­berg Ini­ti­a­tive as well as Char­ity Nav­i­ga­tor.

  • Con­ver­sion: For the peo­ple who have heard of GiveWell, how likely is the per­son to go through GiveWell’s recom­men­da­tions when con­sid­er­ing a dona­tion, and how likely are they to ac­tu­ally fol­low through? This to me seems like the biggest filter, since for many peo­ple, GiveWell’s philos­o­phy and ar­eas of fo­cus just don’t res­onate. They just aren’t in­ter­ested in the causes GiveWell’s top char­i­ties are in. In­ter­est­ingly, this in­cludes many in the EA com­mu­nity (such as those who find an­i­mal welfare or ex­is­ten­tial risk way more im­por­tant) plus more con­ven­tional donors with their own causes of in­ter­est (en­vi­ron­men­tal­ists, donors who pre­fer keep­ing money lo­cal, etc.) GiveWell is not an im­pulse buy. GiveWell doesn’t move much money from peo­ple who ca­su­ally Google for where to donate money, land on GiveWell’s site, and im­me­di­ately buy. Fun­da­men­tally, since GiveWell is ask­ing peo­ple to donate to char­i­ties that they prob­a­bly hadn’t heard of and weren’t con­sid­er­ing ex­plic­itly, it takes some time to get them to ac­tu­ally donate. It also takes time to un­der­stand the broad philos­o­phy mo­ti­vat­ing GiveWell’s de­ci­sions. This is in con­trast with, say, Char­ity Nav­i­ga­tor, that you might go to in or­der to check out a char­ity you are already in­ter­ested in, and then, af­ter read­ing the re­view, de­cide to donate or not donate (it’s hard to say whether Char­ity Nav­i­ga­tor caused the dona­tion, but it did in­fluence the dona­tion de­ci­sion some­what). For more on a com­par­i­son be­tween the im­me­di­ate ap­peal of GiveWell and Char­ity Nav­i­ga­tor, see my Face­book thread dis­cussing the re­sults of a Sur­veyMon­key Au­di­ence sur­vey.

  • Re­ten­tion and up­sell: For the peo­ple who are already donat­ing to GiveWell, how likely are they to re­turn next year, and will the amount they donate keep go­ing up? While this partly de­pends on how good they find GiveWell, it also de­pends on how much money they’re mak­ing. If GiveWell cap­tures po­ten­tial donors when they are young and start­ing their ca­reers, its donor base will see an in­crease in per cap­ita in­come and there­fore will prob­a­bly move larger sums of money per donor.

At the be­gin­ning of my in­ves­ti­ga­tion, I be­lieved that the fo­cus should be on dis­cov­ery and con­ver­sion. This partly stems from my pro-web bias, where short-run pageviews are both the most im­por­tant and fairly de­cent prox­ies for long-run pageviews. How­ever, af­ter I dug deeper into the data, I now be­lieve that the main driver of GiveWell’s growth in the near term is re­ten­tion and up­sell. My up­shot:

  • Dis­cov­ery and brand aware­ness: GiveWell still has a long ways to go be­fore peo­ple have vaguely heard of it and are vaguely aware of it. How­ever, progress in re­cent years has been luke­warm at best. A lot of GiveWell’s me­dia cov­er­age to date was back in 2008 in the wake of the Me­taFilter scan­dal (!), and even in re­cent years, GiveWell re­ceived more at­ten­tion on some met­rics in 2011 and 2012 than it does now. With that said, there seems to have been one dis­crete but per­ma­nent bump in July 2014. 2015 ap­pears to be mostly flat along some met­rics, but ris­ing along oth­ers, paint­ing a de­cid­edly mixed pic­ture.

  • Con­ver­sion: I be­lieve 2015 will see higher con­ver­sion than pre­vi­ous years be­cause of more care­ful co­or­di­na­tion among GiveWell al­lies to move money to GiveWell-recom­mended char­i­ties, plus other rea­sons I dis­cuss be­low.

  • Re­ten­tion and up­sell: GiveWell seems to have suc­ceeded at the up­sell in 2014. This seems to me to be largely a re­sult of GiveWell donors get­ting richer (and there­fore mov­ing up the size buck­ets), but it could also be due to other fac­tors such as in­creased con­fi­dence in GiveWell recom­men­da­tions. Partly, this is also a sig­nal of GiveWell not do­ing that well at at­tract­ing newer donors.

Dis­cov­ery and brand awareness

While we don’t have money moved or GiveWell web traf­fic data from Au­gust 2015 on­ward, we do have other pieces of data that we can use to un­der­stand GiveWell’s growth.

I com­bine pageview counts for the Wikipe­dia page about GiveWell with GiveWell’s spread­sheet with data on the num­ber of unique vis­i­tors. Due to month-to-month vari­a­tion in traf­fic I look at year-over-year growth for a three-month to­tal end­ing at the cur­rent month. So the first row is the per­centage in­crease be­tween the Septem­ber to Novem­ber 2014 pe­riod and the Septem­ber to Novem­ber 2015 pe­riod.

Month Views of page GiveWell Year-over-year growth (three-month to­tal end­ing at the cur­rent month) Unique vis­i­tors (GA) Year-over-year growth (three month to­tal end­ing at the cur­rent month)
201511 2126 −3.14%
201510 1783 −5.75%
201509 1735 3.40%
201508 1840 31.09% 58,371 14.51%
201507 1929 53.40% 58,997 66.47%
201506 1966 60.89% 69,496 58.21%
201505 1695 44.61% 91,280 41.86%
201504 1893 49.77% 69,479 12.29%
201503 1722 60.99% 60,762 6.29%
201502 1660 37.24% 51,995 10.26%
201501 1884 25.99% 67,305 −5.05%
201412 2316 18.10% 145,927 −11.95%
201411 1928 41.12% 68,231 −24.74%
201410 1910 45.40% 64,799 −8.76%
201409 1989 46.04% 56,306 −9.58%
201408 1786 8.53% 76,064 −15.90%
201407 1548 −12.95% 39,238 −27.20%
201406 1041 −20.33% 47,890 −16.99%
201405 1055 −16.96% 44,894 −3.81%
201404 1356 −26.09% 52,754 −1.19%
201403 1261 −38.41% 58,512 −4.87%
201402 905 −32.33% 51,031 18.58%
201401 1105 −25.41% 59,859 49.38%
201312 2260 −17.50% 129,662 75.63%
201311 1499 −13.02% 106,909 89.65%
201310 1452 −2.23% 80,231 101.46%
201309 1178 2.79% 64,443 104.32%
201308 1280 10.59% 71,414 118.98%
201307 1187 22.54% 53,943 92.31%
201306 1564 56.99% 68,686 82.39%
201305 1435 100.64% 58,711 33.86%
201304 1334 169.36% 47,934 32.52%
201303 1653 188.64% 55,700 36.17%
201302 1778 208.71% 60,624 35.46%
201301 1880 179.39% 61,754 4.82%
201212 2652 167.06% 80,480 −8.85%
201211 1989 55.28% 56,205 −21.56%
201210 1675 45.68% 43,695 −17.82%
201209 1083 40.44% 32,758 −20.52%
201208 1241 156.33% 30,810 −13.96%
201207 1222 154.74% 29,325 −3.45%
201206 1182 82.06% 28,476 −7.37%
201205 1012 23.20% 36,494 −4.26%
201204 566 3.39% 31,160 13.80%
201203 626 17.42% 53,627 44.43%
201202 577 47.26% 39,163 110.29%
201201 637 64.83% 37,986 144.78%
201112 830 84.48% 72,603 180.97%
201111 867 167.22% 78,724 249.98%
201110 668 137.05% 46,566 277.10%
201109 1522 133.58% 43,824 308.39%
201108 555 45.10% 40,133 300.78%
201107 448 43.58% 32,925 235.28%
201106 419 66.59% 29,934 206.37%
201105 474 101.24% 34,801 231.79%
201104 623 165.27% 39,044 186.15%
201103 692 58.76% 52,837 82.94%
201102 396 42.95% 17,039 18.61%
201101 479 47.04% 20,671 24.87%
201012 513 124.13% 33,503 47.92%
201011 424 134.43% 23,167 58.11%
201010 345 119.73% 13,763 53.78%
201009 375 89.32% 11,391 50.04%
201008 438 64.15% 9,458 49.54%
201007 268 62.43% 7,771 85.33%
201006 274 58.26% 8,469 118.57%
201005 392 24.16% 12,888 68.58%
201004 244 −12.84% 12,517 69.78%
201003 253 −37.61% 12,776 136.14%
201002 148 −62.03% 12,771 300.57%
201001 586 −74.00% 23,948 435.80%
200912 237 −79.56% 23,319 442.92%
200911 140 −70.74% 14,672 592.54%
200910 195 6.90% 9,623 479.03%
200909 153 39.27% 6,267 489.83%
200908 179 36.61% 6,617 442.80%
200907 239 10.36% 6,191 264.66%
200906 179 −4.96% 4,377 198.79%
200905 157 28.09% 5,149 205.76%
200904 239 50.41% 5,972 112.69%
200903 320 98.74% 11,527 −8.44%
200902 181 292.18% 4,921 −78.52%
200901 1081 4,512
200812 1295 5,555
200811 1328 1,493
200810 175 1,722
200809 165 1,198
200808 153 967
200807 92 1,069
200806 192 1,130
200805 237 2,111
200804 176 1,946
200803 146 3,350
200802 170 5,245
200801 480 14,298
200712 2 50,240

If you dig into the Wikipe­dia pageviews more at the link, you’ll see there was a sharp in­crease in the lat­ter half of July 2014. In fact, the date range for the in­crease can be nar­rowed to the pe­riod be­tween July 11 and Au­gust 4. You can also check against stats.grok.se data.

For this rea­son, we see that three-month pe­ri­ods whose end dates are af­ter Au­gust 2014 but be­fore Au­gust 2015 show ro­bust year-over-year growth, whereas three-month pe­ri­ods Septem­ber 2015 on­ward show ane­mic growth in Wikipe­dia views.

On these in­di­ca­tors, 2015 seems to be a year of flatlin­ing. How­ever, Google Trends data sug­gests ro­bust year-on-year growth from 2014 to 2015, up to and in­clud­ing Novem­ber and De­cem­ber. I’m not re­ally sure how much rel­a­tive weight to place on Wikipe­dia view data ver­sus Google Trends data. Since Google Trends method­ol­ogy is less trans­par­ent, I’m a lit­tle more sus­pi­cious of it. But on the other hand, Google clearly has much more money and brains be­hind the qual­ity of the data. Over­all I would say the ev­i­dence for growth in 2015 (over and above what­ever hap­pened in the mid­dle of 2014) is de­cid­edly mixed. Either way, it is un­likely to be the driver of an in­crease in dona­tions. If dis­cov­ery and brand aware­ness were the only thing that I were look­ing at, I wouldn’t es­ti­mate more than 10% in­crease in an­nual money moved. But I ac­tu­ally es­ti­mate a much greater in­crease (over 40%, with 20% be­ing due to more donors and 20% be­ing due to more money per donor). Why? Ba­si­cally, con­ver­sion and up­sell. Since these two are some­what hard to sep­a­rate, I’ll dis­cuss some changes in 2015 that I be­lieve will pos­i­tively af­fect both con­ver­sion and up­sell, with­out sep­a­rat­ing con­ver­sion from up­sell.

I should also men­tion here some cur­rently fledgeling efforts be­ing made to in­crease brand aware­ness of GiveWell and the ideas sur­round­ing it. Gleb Tsipursky wrote a Huffing­ton Post lis­ti­cle on the 8 Se­crets of Savvy Donors, and he re­ported in an EA Fo­rum post that it in­creased aware­ness and also con­verted at least one mod­er­ately in­fluen­tial per­son. It’s too early to eval­u­ate the sig­nifi­cance of these for the 2015 Giv­ing Sea­son.

The role played by the speci­fics of GiveWell’s recom­men­da­tions and the rhetoric they use

There is a rea­son­able ar­gu­ment that the ac­tual list of top char­i­ties, and the com­men­tary that GiveWell makes when recom­mend­ing them, it­self af­fects how much money is moved to them. The top char­ity list hasn’t changed since the 2014 giv­ing sea­son. But there are a few im­por­tant differ­ences:

  • For Fe­bru­ary-Novem­ber 2014 (the first ten months of the year), the top char­ity list didn’t in­clude the Against Malaria Foun­da­tion. (GiveWell an­nounced in Oc­to­ber 2014 that AMF had es­tab­lished room for more fund­ing again, how­ever it offi­cially up­dated its list only at the end of Novem­ber). Ce­teris paribus, it seems that giv­ing AMF more promi­nence in recom­men­da­tions in­creases money moved, be­cause AMF seems nat­u­rally more ap­peal­ing to GiveWell donors (it’s do­ing some­thing that saves lives ap­ply­ing a proven in­ter­ven­tion). AMF has been in GiveWell’s list for all of 2015, so it’s had more time in the limelight.

  • In its top char­ity an­nounce­ment in 2015, GiveWell has ex­plic­itly dou­bled down on AMF, ask­ing donors to di­rect all their money to AMF. This clar­ity of mes­sage is likely to in­spire unattached donors to give more. Note that GiveWell finds room for more fund­ing in all its top picks, so those de­ter­mined to give to an­other char­ity will not be dis­suaded from giv­ing en­tirely.

  • GiveWell has iden­ti­fied larger amounts of room for more fund­ing in all its top char­i­ties rel­a­tive to pre­vi­ous years, and Good Ven­tures has made sub­stan­tially larger grants this year than in pre­vi­ous years. This recom­men­da­tion an­chors in­di­vi­d­ual donors to give more and makes them more likely to donate, as long as they feel some­what pos­i­tively about at least one of the top char­i­ties [On the flip side, it may make in­di­vi­d­ual donors more re­luc­tant to fund some­thing that already has so much money be­hind it. How­ever, as an em­piri­cal mat­ter, this seems a rel­a­tively minor con­cern among GiveWell donors]. This will most sig­nifi­cantly af­fect up­sell rather than con­ver­sion, but as noted ear­lier, up­sell is a big­ger short-run source of in­crease in money moved than con­ver­sion of new donors.

In­deed, as de­scribed in the 2015 staff mem­bers’ per­sonal dona­tions blog post, most of GiveWell’s staff are donat­ing all or most of their own an­nual money to AMF, with Holden and Elie in par­tic­u­lar donat­ing all their money to AMF. In con­trast, in pre­vi­ous years, GiveWell’s own staff mem­bers were more skep­ti­cal of the rank­ing or rel­a­tive al­lo­ca­tion sug­gested by GiveWell, and opted for a differ­ent mix. GiveWell’s staff also ex­pressed more con­fi­dence in 2015 (jus­tified or not) in the rigor of the pro­cess used to eval­u­ate top char­i­ties and their room for more fund­ing. For in­stance, co-founder and co-ex­ec­u­tive di­rec­tor Elie wrote:

Last year, we gave sig­nifi­cantly more to GiveDirectly than GiveWell recom­mended be­cause I put less weight on cost-effec­tive­ness anal­y­sis than other staff. I re­main am­biva­lent about putting so much weight on cost-effec­tive­ness es­ti­mates be­cause they are sub­ject to nu­mer­ous de­bat­able as­sump­tions, but I’ve changed my mind based on work we did over the past few months. We put sig­nifi­cant effort into de­bat­ing our cost-effec­tive­ness model (.xls) and en­couraged all GiveWell staff to in­put their own val­ues into the model. The re­sult­ing strong con­sen­sus that cash is sig­nifi­cantly less cost-effec­tive than bed­nets/​de­worm­ing per­suaded me to put more weight on cost-effec­tive­ness anal­y­sis in my giv­ing this year.

Similarly, Alexan­der Berger notes his in­creased con­fi­dence in the recom­men­da­tions and how much AMF strikes him as a great op­por­tu­nity:

What I wrote last year is even more true this year: I think our recom­men­da­tions this year are stronger than they were last year, and I’m ex­cited to sup­port them. And since I’ve been (even) less in­volved in our top char­i­ties re­search this year than in pre­vi­ous years, I’m also in­clined to just fol­low the recom­mended GiveWell al­lo­ca­tion (i.e. 100% to AMF).

I’m par­tic­u­larly struck this giv­ing sea­son by just how strong a recom­men­da­tion AMF is. We be­lieve that AMF can fairly re­li­ably fill a gap in global bed­net cov­er­age and avert child deaths for some­thing on the or­der of ev­ery $3,000 donated. I’m much more skep­ti­cal than some of my col­leagues about how much we can learn from our cost-effec­tive­ness mod­els, par­tic­u­larly for de­worm­ing, but I see the case for bed­nets as be­ing quite strong. And even if we are off by a fac­tor of sev­eral times, which I think is ab­solutely pos­si­ble, I see that as re­mark­ably cost-effec­tive: for com­par­i­son, com­mon es­ti­mates put the value of a statis­ti­cal life in the United States at around $5 mil­lion.

In­creased co­or­di­nated dou­bling down on GiveWell recom­men­da­tions by its allies

One ma­jor thing I see be­ing done at a larger scale in 2015 is co­or­di­na­tion by or­ga­ni­za­tions that seek to move money to GiveWell top char­i­ties through care­ful mar­ket­ing and mes­sag­ing. GiveWell, which has his­tor­i­cally re­mained very de­tached from efforts to spread the word about its top char­i­ties, has given a more ex­plicit nod to fundrais­ing efforts ex­plic­itly al­igned with the goal of rais­ing money for its top recom­mended char­ity. For in­stance, their Outreach As­so­ci­ate Cather­ine Hol­lan­der (the po­si­tion didn’t even ex­ist last year) wrote a blog post re­cently on fundrais­ing for AMF, the top char­ity. Char­ity Science’s fundrais­ing op­er­a­tion also ap­pears to have ma­tured this year, as ex­em­plified in their post pro­vid­ing guidance for Christ­mas fundraisers. Giv­ing What We Can also seems to have upped its so­cial me­dia pres­ence and the pub­lic­ity sur­round­ing its Pledges. While not all the money moved through this will be at­tributed to GiveWell, it may raise the pro­file of GiveWell and move some money ex­plic­itly at­tributed to GiveWell.

More on the dis­tri­bu­tion of money moved to top charities

In 2014, AMF was not in the top char­ity list till the end of the year, but still raised $4.4 mil­lion from non-GV donors, higher than any other char­ity. This year, AMF was a top char­ity through­out the year, and GiveWell iden­ti­fied as the sole char­ity that it recom­mended GiveWell-al­igned donors to give to. GiveWell it­self, and its al­lies, are pick­ing on AMF as the fo­cal point of fundrais­ing cam­paigns. AMF is also in­trin­si­cally ap­peal­ing to GiveWell donors, since it’s a scal­able im­ple­men­ta­tion of a proven in­ter­ven­tion that saves lives. For all these rea­sons, I be­lieve that AMF will raise the most money among GiveWell-recom­mended char­i­ties by a huge mar­gin. More­over, the un­cer­tainty around money moved to AMF ac­counts for most of the un­cer­tainty in to­tal money moved.

I be­lieve the other char­i­ties will gen­er­ally re­ceive money from peo­ple who have strong prior rea­sons for want­ing to donate to the char­i­ties, with the GiveWell recom­men­da­tion play­ing a Char­ity Nav­i­ga­tor-like role of val­i­dat­ing the or­ga­ni­za­tion’s le­gi­t­i­macy. For in­stance, GiveDirectly ex­cites a large num­ber of effec­tive al­tru­ists, due to the sim­plic­ity, non-pa­ter­nal­ism, strong or­ga­ni­za­tional growth, and sig­nifi­cant scal­a­bil­ity. Peo­ple with whom these val­ues res­onate will likely con­tinue to donate to GiveDirectly and use the huge fund­ing gap GiveWell has found for the or­ga­ni­za­tion to bolster their case. Similarly for SCI and DtWI. It’s also im­por­tant to note that since the dou­bling down on AMF hap­pened in Novem­ber, the first few months prob­a­bly saw a larger share of dona­tions to these other char­i­ties.

Growth in com­ing years

Most of my dis­cus­sion so far has been fo­cused ei­ther on gen­eral mod­els for money moved or what’ll hap­pen in 2015. How­ever, my pri­mary in­ter­est is in un­der­stand­ing growth over the next 5 or more years. The rea­son I de­voted so many pix­els to 2015 is that the state­ments I make about 2015 will be tested against re­al­ity in fairly short or­der, and I be­lieve this cal­ibra­tion against re­al­ity will be cru­cial to eval­u­at­ing my cred­i­bil­ity.

Let’s get back to three de­ter­mi­nants of GiveWell’s money moved: dis­cov­ery and brand aware­ness, con­ver­sion, and re­ten­tion/​up­sell.

  • Dis­cov­ery and brand aware­ness: I be­lieve that this is low, but I don’t be­lieve that in­creas­ing it right now will do much. I think in­creas­ing brand aware­ness will do very lit­tle till the deeper prob­lem of low con­ver­sion is solved. As the sur­veys I did showed, GiveWell brand aware­ness is low but not that low. It’s par­tic­u­larly high among peo­ple who already use Char­ity Nav­i­ga­tor, and yet, most of the peo­ple who are aware of GiveWell don’t ac­tively use it.

  • Con­ver­sion: 2015 might already be a peak in terms of the ex­tent to which the char­i­ties con­vert peo­ple. I think that, over the timeframe of the next two years, there’s a greater chance that AMF, SCI, DtWI, or GiveDirectly would run out of room for more fund­ing as judged by GiveWell than that GiveWell will find an­other promis­ing char­ity to which it recom­mends mov­ing a com­pa­rable amount of money. I also be­lieve that fundraiser co­or­di­na­tion is also reach­ing close to the best it can achieve in squeez­ing out money. I think we are already near­ing the long-run bot­tle­neck on con­ver­sion: GiveWell char­i­ties just don’t res­onate with enough peo­ple.

  • Re­ten­tion and up­sell: There is a mo­men­tum effect here aris­ing from the fact that GiveWell got a large num­ber of young donors in its early years. The in­come and giv­ing po­ten­tial of GiveWell donors will con­tinue to in­crease over the next few years, so that alone will be a driver of growth. I don’t be­lieve this will run out for the next 2-3 years, but it could run out af­ter that if GiveWell doesn’t add enough new donors with high earn­ing po­ten­tial in the pipeline.

The up­shot: most of the growth in money moved to GiveWell top char­i­ties in the next two years will come from re­ten­tion and up­sell. More con­cretely, this means that the num­ber of donors prob­a­bly won’t in­crease that much from 2015 to 2016 and 2017, but the money moved per donor will con­tinue to in­crease, largely be­cause of an in­crease in donors in high size buck­ets (and mostly this will be donors who donated smaller amounts in ear­lier years). More con­cretely, I ex­pect the num­ber of donors to in­crease by 10% per year for the next two years (2015 to 2016 and 2016 to 2017) and I ex­pect the money moved per donor to in­crease by about 20% and 15% in the com­ing two years. A lot of this will come from the high­est dona­tion size buck­ets, via phe­nom­ena such as Founders Pledge dona­tions.

I don’t think the time is right to spell out the model in de­tail. I’d like to see how close my 2015 pre­dic­tions are to re­al­ity and un­cover any bi­ases in how I’m think­ing about things be­fore elu­ci­dat­ing a plau­si­ble story for the next few years.

Implications

To the ex­tent that I be­lieve the me­dian case of my fore­casts, it leaves me with mixed feel­ings. I be­lieve that there is wis­dom in hav­ing larger num­bers of eye­balls en­gag­ing with the data and recom­men­da­tions. The idea of growth be­ing pro­pel­led more by re­ten­tion than by user growth is a nega­tive as far as qual­ity is con­cerned. On the other hand, as­sum­ing some ra­tio­nal­ity on the part of donors, larger sums of money moved per donor in­cen­tivizes each donor to care more about their dona­tions, and could in prin­ci­ple put more scrutiny on GiveWell. I be­lieve that the former effect sig­nifi­cantly out­weighs the lat­ter at the pre­sent mar­gin, par­tic­u­larly be­cause a lot of the growth in money moved per donor comes from donors just get­ting richer, rather than de­cid­ing to donate a larger frac­tion of their in­come to char­ity.

I also think the con­tinued short-run in­crease in money moved to GiveWell top char­i­ties re­duces the in­cen­tives for GiveWell to ex­pand to more do­mains and un­der­take the hard task of find­ing new ar­eas to donate to. This limits their long-term ap­peal (in par­tic­u­lar, it puts a ceiling on how much con­ver­sion they can ac­com­plish), but more im­por­tantly from a so­cial welfare per­spec­tive, it limits the frac­tion of the donor pop­u­la­tion and the frac­tion of donor money that will be in­fluenced by the kind of deep anal­y­sis that GiveWell strives to do. (With that said, Char­ity Nav­i­ga­tor-level or GuideS­tar-level anal­y­sis of char­i­ties has taken off in a de­cent way and is prob­a­bly in­form­ing over 1% of money moved already, but that’s not the same type of anal­y­sis since it’s based on a much smaller, though still im­por­tant, set of met­rics).

UPDATE 1: In a blog post com­ment on De­cem­ber 23, Elie Hassen­feld re­ported that as of De­cem­ber 22, GiveWell had moved $15.7 mil­lion in non-Good Ven­tures dona­tions to its top char­i­ties. My origi­nal es­ti­mates did not in­clude this in­for­ma­tion, but ac­count­ing for it would move the lower end of my es­ti­mates up­wards some­what while not sig­nifi­cantly af­fect­ing the higher ends.

UPDATE 2: In a blog post on Jan­uary 8, 2016, Tyler Heish­man re­ports GiveWell’s Q3 2015 money moved and also pro­vides a pre­limi­nary es­ti­mate of the money moved be­tween Fe­bru­ary and De­cem­ber 2015. This doesn’t in­clude money that might be moved in Jan­uary 2016, though that is un­likely to ma­te­ri­ally af­fect the es­ti­mates much since the bulk of money moved is in De­cem­ber. Heish­man re­ports that $28 mil­lion in money moved from sources other than Good Ven­tures is at­tributable to GiveWell. This is close to the 90th per­centile in my es­ti­mate range ($28.6 mil­lion). So over­all, my es­ti­mates were quite con­ser­va­tive, and it may have been pos­si­ble to more ac­cu­rately make higher es­ti­mates. With that said, my iden­ti­fi­ca­tion of the po­ten­tial rea­sons for the money moved be­ing a lot higher than my me­dian es­ti­mate was broadly cor­rect. As I wrote above:

The up­per end of es­ti­mates is mainly gov­erned by fac­tors such as wealthy donors, highly suc­cess­ful fundrais­ing groups or celebrity en­dorse­ments or me­dia pub­lic­ity, and other un­fore­seen fac­tors.

The GiveWell blog post says:

Ex­clud­ing Good Ven­tures, we have tracked about $28 mil­lion (of which, roughly half has come from donors giv­ing $1 mil­lion or more).

Hence, at least the “wealthy donors” part seems right.

More­over, I ex­pect that the other pre­dic­tions I made should roughly hold up with ap­pro­pri­ate rescal­ing to the greater amount of money moved, but also ac­count­ing for the in­creased vari­abil­ity due to a small num­ber of large donors.

UPDATE 3: Cather­ine Hol­lan­der linked to al­most-fi­nal es­ti­mates of money moved by char­ity and size bucket. The to­tal non-Good Ven­tures money moved was about $33.37 mil­lion (ver­sus about 13 or 14 mil­lion last year, and a 50% con­fi­dence in­ter­val of $16.2-24.5 mil­lion in my post). Within the dis­tri­bu­tion speci­fied in my post, the amount ac­tu­ally moved fell at the 97.5th per­centile.

Size buck­ets: Money from those giv­ing over $100,000 (in­clud­ing both the $100,000-$999,999 bucket and the $1,000,000+ bucket) was about $20.29 mil­lion, ver­sus $4.42 mil­lion last year. Ex­clud­ing these heavy donors, the money moved was $13.4 mil­lion, com­pared to about $8.4-10 mil­lion last year. The money moved per donor was $2,157.

I be­lieve my pri­mary source of er­ror in es­ti­mates was at the high end. I cor­rectly pre­dicted that the money moved per donor would in­crease, but I un­der­es­ti­mated the ex­tent to which the to­tal money moved by donors giv­ing in ex­cess of $100,000 would in­crease. If the giv­ing in these buck­ets had in­creased in the same pro­por­tion as the giv­ing in the sub-$100,000 buck­ets, the money moved would have been within my 50% in­ter­val. With that said, I also un­der­es­ti­mated some­what the sheer vol­ume of donors: my 50% es­ti­mate for the num­ber of donors was 9,500-12,500, whereas the ac­tual num­ber was 15,468.

Dona­tions by re­cip­i­ent char­ity: Here, too, my es­ti­mates were off, but not sys­tem­at­i­cally so. Three of the four top char­i­ties (Against Malaria Foun­da­tion, Schis­to­so­mi­a­sis Con­trol Ini­ti­a­tive, and De­worm the World Ini­ti­a­tive) fell in my 50% in­ter­vals for those char­i­ties (with AMF, at al­most 15 mil­lion, be­ing at the higher end of its range but still within it). The one out­lier was GiveDirectly. My 50% es­ti­mate for GiveDirectly was be­tween 1 and 4 mil­lion dol­lars, but ac­tual money moved ap­pears to have been over 13 mil­lion dol­lars. This is in­ter­est­ing in that it was un­ex­pected not only to me but also (prob­a­bly) to many other GiveWell watch­ers. It was also con­trary to the mes­sage that GiveWell was fo­cused on spread­ing, namely, that donat­ing to AMF should be the top pri­or­ity.

UPDATE 4: Natalie Stone-Crispin of GiveWell has just pub­lished a blog post on the GiveWell blog pro­vid­ing in­sight into GiveWell’s re­search plans for 2016. Some of these up­dates are re­lated to pre­dic­tions I made, and make me more con­fi­dent that my think­ing was in the right di­rec­tion. In my Face­book post that had been the origi­nal im­pe­tus for this Effec­tive Altru­ists group post, I had said that GiveWell’s top char­i­ties are un­likely to change in 2016, and Natalie’s just-pub­lished post strength­ens my view that the top char­ity list is likely to stay the same (with pos­si­ble tem­po­rary re­moval of some top char­i­ties be­cause of ex­haus­tion of room for more fund­ing or con­cerns about scal­ing challenges). GiveWell is also look­ing into switch­ing fo­cus from iden­ti­fy­ing top char­i­ties to out­reach, some­thing that I had not made a pre­dic­tion on in ei­ther di­rec­tion.

UPDATE 5: On Fri­day, May 13, 2016, GiveWell fi­nally re­leased the fi­nal re­port on money moved in 2015. I have pub­lished a new post go­ing over my fore­casts, how they com­pared with re­al­ity, and im­pli­ca­tions.

Dis­clo­sure: I am not fi­nan­cially af­fili­ated with GiveWell, but have some past and on­go­ing as­so­ci­a­tion with them. You can get a de­tailed ac­count of my past in­ter­ac­tion and af­fili­a­tion with GiveWell here.