Pablo Melchor, co-founder and president at Ayuda Efectiva. Co-founded EA Spain.
There is no perfect calculation of all the effects of a program but I think GiveWell's effort is impressive (and, as far as I can tell, unmatched in terms of rigor). I think the highest value is in the ability to differentiate top programs from the rest, even if the figures are imperfect.
Hi Aaron,I think it's great that you ask these questions. I wouldn't assume that the majority of the community already has a crystal-clear grasp of them since (1) they are not straightforward at all and (2) as far as I know, the answers are not really consolidated in some single post you can read in 5 minutes.
GiveWell's current estimate is $3,000-$5,000 per life saved. This is the range they communicate on their Top Charities page and which they explain here. They have not updated this messaging since November 2020, so that may change soon.
As for a rough overview of the calculation process, this example may help. It is a complex process that starts with the evidence of effectiveness for a particular intervention but then includes a host of factors for which GiveWell calculate and regularly update their best estimates. Some mentioned in the example I just linked to are:
If you want to dig deeper, you can go over the cost-effectiveness spreadsheets on this page Michael shared in a previous answer or read this detailed guide.
Your second question was about what "saving a life" actually means. Holden (GiveWell co-founder, now Open Philanthropy's co-CEO) wrote this post about it in 2007. Some snippets:
Approximately 80% of children born in sub-Saharan Africa reach age 5; in the developed world, it’s 99%. [...]Between ages 5 and 45, people in sub-Saharan Africa have relatively similar mortaility[sic] rates to those in the developed world except for the influence of HIV/AIDS, TB, and mothers dying in childbirth. Around age 45, a lot of the same diseases that kill people under 5 start killing again (maybe due to weakened immune systems). [...]So what’s a life saved? If you save someone right as they exit infancy (5 years old), you’ve saved someone who probably has around a 50% chance of making it to age 60 … another way of putting this is that if you save two lives [...], you’ve in expectation given one person a full life that they wouldn’t have had.
Approximately 80% of children born in sub-Saharan Africa reach age 5; in the developed world, it’s 99%. [...]
Between ages 5 and 45, people in sub-Saharan Africa have relatively similar mortaility[sic] rates to those in the developed world except for the influence of HIV/AIDS, TB, and mothers dying in childbirth. Around age 45, a lot of the same diseases that kill people under 5 start killing again (maybe due to weakened immune systems). [...]
So what’s a life saved? If you save someone right as they exit infancy (5 years old), you’ve saved someone who probably has around a 50% chance of making it to age 60 … another way of putting this is that if you save two lives [...], you’ve in expectation given one person a full life that they wouldn’t have had.
I will find out whether GiveWell has revisited that more recently.
I see this tag is largely unused, while "impact assessment" seems to have been chosen for many posts that could fit here (e.g. https://forum.effectivealtruism.org/posts/ZeFcfCAncT3jPAeht/clean-technology-innovation-as-the-most-cost-effective).
My experience in business matches two of the points that Catherine makes above:
My guess is that it is probably better to have a not-perfect name that everyone uses, than a whole variety of different names.
Another cost is that there are people who hear "effective altruism" several times in several places before deciding to learn more/ get involved, so each exposure of that name (as long as it is positive!) helps.
My current view is that:
If anything, I would say that one of the weaknesses of Effective Altruism (purely from a branding perspective) is that its brand landscape is already super-diverse (e.g. there is GiveWell, ACE, Open Phil, Founders Pledge, GWWC, 80,000 Hours, etc., etc. each pushing their own brand). This does make sense since each of the organizations I mention is applying effective altruism to a particular space or situation. However, when it comes to local groups, I tend to think that the EA movement as a whole has much more to gain from consistency.
Ben mentions in his comment how independent brands can reduce brand risk for EA, which is true. However, I think they can also reduce brand potential for EA (this is more of a side note, but I think that whenever we consider minimizing reputational risks we should also consider the opportunity costs of not doing something or doing it in the cautious-but-probably-less-impactful version).
I think that if we want to make the EA brand better (more meaningful, attractive, easily recognizable, etc.), simply using it consistently will go a long way.
I checked with our tax advisors and the situation is more or less what I described:
It seems that regranting —as we are doing in Ayuda Efectiva— is the safe way to go for now.
Hi Jan-Willem, thanks for the info!I am interested in learning more about DonerEffectief: I will DM you about that.As for EU-wide deductibility, last time I checked it was one of those cases where:
I am quite sure that is still the case in Spain, but I will use this as a nudge to look into it again :-)
Devon, thanks for the comment!
On that topic, do you or anyone else know of other such national level projects, aside from Effektiv Spenden (where I do some work) in Germany and RC Forward in Canada?
I know of gieffektivt.no in Norway and Effective Altruism Australia.
Although I imagine there are many nuances of each country's bureaucracy, it could make sense to template out the operational aspects that likely don't differ from country to country.
Certainly. I had a brief conversation with Basti and Jack Lewars on this. I am very skeptical about large and complex projects like migrating to the same ERP but would be happy to explore more modest opportunities to avoid repeatedly reinventing the wheel :-)
What kind of fixed costs are we talking about, if you don't mind sharing? If they were too low, I'd be a bit concerned that it could hamper your multiplier. How are you approaching fundraising?
Fair concern and you are right: "the lowest possible fixed costs" can mean anything :-). In our case, we are looking at costs of less than 10K €/month for 2021. The plan is to bootstrap this (which we can afford) until we can show impact. The metric we will be focusing on is what The Life You Can Save calls leverage ratio: money moved expressed as a multiple of operating expenses. If you are interested in getting into the weeds, I'd be happy to chat :-)
I made a minor edit to clarify a couple of points in "How you can help".
Yes, thanks for that: I can see the broader strategic implications. I actually think the equivalent to "but actually may lead to more people joining top priority paths in the focus areas of existing career orgs in the long run" may also be true in the effective giving space.
Sella, thanks for the post. I think this is a very interesting idea (and I am guessing that other non-US/UK EA groups may think so as well). I see it as doing relative optimization in a much larger space rather than absolute optimization within a small group (people who actually have a chance of going into 80,000 Hours's highest-impact paths).
In that sense, Probably Good reminds me of what Elijah explained here about what the ImpactMatters team is trying to do under their new roof at Charity Navigator:
Certainly in typical EA terms, many of the nonprofits that are analyzed are not the most cost-effective. But we also know that standard EA nonprofits are a fraction of the $300 bil nonprofit sector, and there is a portion of that money that has high intra-cause elasticity but low inter-cause elasticity. Impact analysis could be a way of shifting that money, yielding very cost-effective returns [...]