Subscribe
Share
Share
Embed
Welcome to GiveWell’s podcast sharing the latest updates on our work. Tune in for conversations with GiveWell staff members discussing current priorities of our Research team and recent developments in the global health landscape.
Elie Hassenfeld: [00:00:00] Hey everyone. I'm Elie Hassenfeld, GiveWell's co-founder and CEO. Today we're going to have a follow-up conversation on the GiveWell story that was recently featured in the November 26th, 2025 episode of NPR's Planet Money podcast.
That episode covered our $2 million grant to an organization called ALIMA. That's an acronym that stands for the Alliance for International Medical Action. But the organization just goes by ALIMA, so that's the word we're gonna use. This is a grant that provided healthcare and nutrition services to children in need in Cameroon, following unexpected aid cuts earlier this year.
If you haven't listened to the episode on Planet Money yet, titled Saving Lives with Fewer Dollars, I recommend you do that before listening to this conversation. That episode provides the context for what we'll discuss today, and you can find it on NPR's Planet Money podcast feed. We'll also link it in the summary for this episode.
In today's conversation, I want to dive into this grant in more detail to discuss the timeline of [00:01:00] our investigation from start to finish, how it was similar and also different from our normal approach, going deeper on some of the modeling questions that we tried to answer. Also, what's happened since we've made the grant and an update on the grant and ALIMA's work in Cameroon.
And I hope that all of this helps illustrate how we as a team at GiveWell responded to this year's urgent needs created by foreign aid cuts. So I'm joined today by GiveWell researchers Rosie Bettle and Alice Redfern. Thank you both for having this conversation with me. Could you each introduce yourselves, your roles at GiveWell, and the role that you each played in the ALIMA grant?
Rosie Bettle: Sure. So my name is Rosie Bettle. I'm a program officer at GiveWell. I'm now on the vector control team, so thinking about malaria. And during the grant investigation, I was working on the rapid response team, working with Alice to try and figure out if we wanted to make this ALIMA grant and to broadly look at were there any other USAID gaps falling between the cracks [00:02:00] of what our other teams focus on that we'd want to investigate.
Alice Redfern: Yeah. And hi everyone, it's Alice Redfern. I'm a program officer for nutrition and the team lead for nutrition. And same as Rosie, on this grant, we were working very much together to figure out whether to make this grant and evaluating other grants in response to the USAID cuts at the time.
Elie Hassenfeld: Thank you both. One of the things that came up a lot in the discussion in the episode was the urgency of ALIMA's need for funding. And the episode threw out some dates at various times as we were going through our investigation, but I found it hard to track just how long this took. And so could you just walk us through the timeline and also if you can, how that timeline in your experience compares to a typical grant investigation at GiveWell.
Rosie Bettle: Sure. So this investigation took us about, I think it was about six weeks from start to finish, from when we first started, you know, having calls with ALIMA to making a final grant decision. And at the [00:03:00] same time me and Alice were also, this wasn't our sole focus. We were also trying to track various other USAID related gaps, and so on.
Elie Hassenfeld: Yeah, six weeks is pretty quick. How does that compare, to the extent you can say or have the experience to, you know, a grant investigation for one about this size, about $2 million.
Alice Redfern: I think it's a reasonable amount of time, but what it misses in most of our grant investigations, there's a lot of pre-work that goes into it.
Normally we have, by the time we're even writing an investigation plan, we have a reasonably good sense of what the program is about, what's going into the grant. There's, you know, months potentially of lead up, if not years into that point. And then we hit the starting on the clock. And then six weeks, I would say is quite reasonable for a grant of that size.
But with this one, what's different is that we really hit the beginning of everything on that first day. We had to go from knowing nothing to fully understanding the program within that timeframe.
And as Rosie mentioned, I think one of the challenges was just that it was not the only thing that we were working on. There were so many moving pieces at that time, for everyone [00:04:00] working in the sector, so that's one of the issues with it moving quickly as well.
Elie Hassenfeld: I mean, for what it's worth, and I guess I'm a biased party here, but for me, six weeks from start to finish sounds pretty fast to go from introduction to a potential opportunity to, you know, being in a position to commit funds and push them out the door.
Just to make sure people understand, because I think people may not have this full context: The way our research process works is before we kick off a formal investigation, we as a team, we go through a formal process of signing off on an investigation plan for a particular grant. And from the post-formal sign off for the investigation plan to a grant to commitment, that's, say, roughly six weeks.
But that investigation plan itself builds on a lot of specific work about a grant opportunity itself, nevermind the program area in general. So often, like thinking of six weeks from start to finish would likely be a lot quicker than a normal investigation that we would do.
I'm curious if you can say, you know, where had we done pre-work for this grant and where had we not done? So, GiveWell has made grants to ALIMA before, we've made grants to ALIMA [00:05:00] for severe acute malnutrition. We've done a lot of work on malnutrition over the course of the last five years.
And so in that way, we were able to do this in six weeks, not literally going from zero to something, because we had all that previous context. But, you know, compared to other investigations that you've done, how did this grant compare in terms of what had been done in the pre-work versus what had to get going when that clock started ticking?
Alice Redfern: Yeah, I'm happy to take this as I've had more experience with nutrition grants since then, which are the most similar. I think that the big thing that was different about this was we had all the pre-work from funding malnutrition programs and from working with ALIMA before. So we had the relationship set up, we knew who we needed to be talking to.
The biggest difference is that whereas normally our malnutrition programs that we're funding are the vast majority malnutrition, this one is the inverse. The majority of the programming that they're doing in Cameroon is not malnutrition, it's basic primary health services. It's being based in the hospital, complementing hospital services across [00:06:00] a very wide range of conditions. And I think that's where me and Rosie probably felt like we were starting a bit more from scratch, was how do you model the benefits of a primary healthcare system? And we were drawing much more on what the work of some of the new areas team has done and just some rough, rough estimates of how to go about that.
Yeah. What do you have to add on that, Rosie?
Rosie Bettle: Yeah, like a hundred percent. Some of the kind of constituent interventions that ALIMA carries out, we have looked at before in some depth, especially CMAM, so treating malnourished kids with ready-to-use therapeutic food. And we've done at least some work understanding malaria treatment, and so on.
But looking across the whole kind of specter of what ALIMA does in this sort of primary healthcare context, you know, doing all these treatments at once, some is inpatient, some is outpatient. And then at the kind of broader level, supporting the healthcare clinics in general through training staff, through [00:07:00] disease surveillance, through buying equipment for the hospitals, all this kind of stuff, meant that the scope felt quite wide for something that we were starting, not entirely afresh, but more afresh than normal for a grant on this kind of timeline.
Elie Hassenfeld: One of the things that the Planet Money episode didn't go into as much as I would've liked is, you know, what ALIMA actually provides, like, what their services are. Could you just say a little bit more about what the funds that we directed to them are actually doing, like what it's actually paying for, and also how that's different from other malnutrition grants that we've made in the past.
You alluded to this a little bit in how it leans more on primary healthcare services versus nutrition-specific services, but would love to get a better handle on both of those if you can share.
Alice Redfern: Yeah, sure. So, a lot of the funding from this grant goes directly to health workers' salaries within the system that they're supporting. I think the [00:08:00] easiest way to think about it is that there's just an insufficient number of healthcare workers to serve the very vast needs in this area. And around two-thirds of the health workforce within the districts served are funded through ALIMA. They're often contracted by the Ministry of Health, but with salaries paid by ALIMA.
ALIMA also brings in a substantial number of their own staff, and they put them in these hospitals as well. So it's a mixture of the two. They also then go out and hire and train directly community health workers, for example.
Another big component is stock. So there's some stock that will come from the Ministry of Health, but there's often stock outs. So ALIMA will procure a buffer. So they might do that for malaria treatment where there's a reasonably good supply chain, but they'll want to have some in stock just in case it runs out.
And then for more complicated treatments, they will often be the only people procuring those medications. Treatments for, you know, neonatal sepsis, very, very sick babies just after birth. And then there's some more kind of complicated things related to, [00:09:00] you know, keeping a hospital clean, running like the sanitation services of a hospital, keeping a neonatal unit open within the hospital. Just kind of operation and logistics cost within the hospitals and the primary care facilities. That's probably the bulk of it, and then some outreach within communities as well.
Elie Hassenfeld: Got it. So this grant is, I know there's like several different things it's doing, but in a large part, it's paying healthcare workers. It's ensuring that needed medicines are in stock. And then it's providing other support that enables hospitals to stay open and function.
On the healthcare workers specifically, what would happen in the absence of philanthropic funding? Were the workers just not being, they wouldn't have been paid so they wouldn't have come to work or, you know, it enabled more patients to be seen? You know, what happened because of philanthropic funding with respect to healthcare workers here?
Alice Redfern: Yeah, often what we heard was you went from a primary healthcare facility that had multiple staff covering shifts, always open, always someone on site, to after the cuts, occasionally someone would be able to [00:10:00] go to that facility and open it for a couple of days a week. So, it's just really cutting down on the amount of time that that care is available.
And then in the hospital it's a bit more complicated because the hospital will obviously rearrange the staff that they have to offer a minimum level of care. So they wouldn't just completely stop treating children, even if all of ALIMA's supported pediatricians disappear. But it becomes very difficult over time for them to maintain the same level of treatment.
Elie Hassenfeld: That reminds me of things that I heard in Malawi when I was there last summer about patients coming to get their HIV treatment. The staff that had been there to deliver the HIV treatment had been laid off, they were gone. And so there was other staff trying to distribute the drugs, but these other staff didn't know how to distribute the drugs. And so they were like trying to match up pill bottles to figure out whether they were giving the patient the right thing.
So there is some staff there to do the work, but they're not in a position to do it in the way that one would, if they had been appropriately or properly trained.
One of the things that the [00:11:00] episode focused so much on was this question of the mortality rate. I think it makes sense in this context where presumably, in a location with extremely high mortality, the need to maintain basic healthcare, via healthcare workers, medicine stock, you know, hospital functioning, some community outreach, that is so necessary in a place with such high mortality.
And so the more that we could get at the question of, well, how high is this baseline mortality, that gives us a lot of the information we needed to figure out whether directing money to this grant, you know, versus all the others that we could have was the right thing to do. I don't know, is that basically right? How would you describe that?
Rosie Bettle: Yeah, I think some like nuance, I mean, Planet Money could not realistically go into for like a 30-minute episode is all the different kinds of mortality that we're trying to understand.
So of course we really wanted to get exactly what you're saying, the baseline mortality. I mean, we were pretty confident that it was very high, there was a question of exactly how high. And [00:12:00] that feeds into our models, in the sense of understanding what kind of reduction in mortality is feasible with this kind of program.
Then the other kinds of mortality effect we're trying to understand, one was whether we could try and get at the effect on mortality of ALIMA's programming as a whole. For example, if they had pre-program results and then post-program results, so we could see that change over time. We did get some of that from some of their other programs. It was unfortunately kind of difficult to interpret since obviously it's not randomized and there's a lot of other things going on in the population at the same time.
And we're also thinking a lot, we would not ask this from ALIMA, this is something we have to work out on our end, like you could not ethically get this data. But we also spent quite a lot of time thinking about the untreated mortality rate from these different conditions that ALIMA is treating.
So kind of like three prongs of mortality questions that we were trying to [00:13:00] understand better.
Elie Hassenfeld: Got it. And so the three prongs are, what is the baseline level of mortality in this region? The second one is, what can we observe about ALIMA's effect on mortality from the data they have, albeit imperfect? And the imperfection there is, you could see, you know, before and after data. But if ALIMA came into an area, presumably there's other things that are happening there, and so it's unclear whether ALIMA alone is the cause of that change in mortality versus other factors. And then finally, if we know that ALIMA's presence is causing certain conditions to be treated that wouldn't otherwise be treated, we could try to add up the mortality as best we can from those different conditions and say, well, these will be treated, so we should expect that overall mortality would fall by some amount.
Rosie Bettle: Exactly, all those three things. And I think each one of them by itself is very challenging. And then try to manage all of those three at once. And we spent a fair bit of challenge, how to just understand, is there really good data somewhere where we [00:14:00] could get a sense of the typical mortality reduction effect of a good humanitarian program?
But it's just crazily difficult to get this because of the point that you are saying, you can't randomize it. You can't quite understand what else is going on to that population beyond that specific humanitarian program. So yeah, these are some of the questions that we were grappling with.
Elie Hassenfeld: We're talking about an area that clearly has such a significant need for support. You know, we know that the mortality rate is high even if we don't know precisely what it is. We know that there was cuts that meant healthcare workers were not getting paid, which logically leads to, you know, much less and worse service.
You know, do you have a sense of the potential variation in the impact ALIMA could have had, from the low end to the high end and, you know, what that would mean for how this program compares to other options?
I mean, I think something that people listening to this probably [00:15:00] recognize is, GiveWell has a significant amount of funding to direct, but it's still very limited relative to the total scope of the need that exists in the world. And so when we look at an opportunity, the question is, should we support ALIMA in Cameroon, or should we support another program somewhere else that could do more? And we really feel that constraint that comes from the resources we have.
But I think it might just be helpful to, to the extent you can, like put some numbers on that low end and high end and what that could have meant for whether or not this was a program that we thought was the right use of the funds that we're responsible for.
Rosie Bettle: Yeah. I think, like, what was interesting about this one is that maybe from the podcast, I could understand people feeling like, oh my gosh, they clearly should have just granted straight away, like super high need area. You know, ALIMA's good, they're doing good stuff, why don't they just like, do it and make the decision very quickly.
But actually from like our perspective, we were pretty uncertain. This was not a given from the start. And I think we really [00:16:00] remained pretty uncertain until fairly close to our final decision.
And some of the reasons for that, I'd highlight that ALIMA, the cost per person treated is a little bit higher than some of our other things that we support, right? It's about like $40, and obviously that is then doing incredibly critical work, you know, you are directly treating people on the ground, but just to give people a sense of scale, we are often supporting interventions which are reaching people far, far cheaper than that, right? Sometimes, $6 for like seasonal malaria chemoprevention, for example.
And trying to weigh up, something like this, it's directly treating people who need treatment, super high burden population, when you could otherwise spend a bit more of that funds on something which is also really critical and you can reach, you know, more people. That for me is the crux of why it was not a given coming in that we would grant to them.
Elie Hassenfeld: And so maybe one way [00:17:00] to conceptualize this quantitatively is to say, very roughly, this program is about 10 times more expensive per person reached than some of the other programs that we support.
And so, it would have to have 10 times more impact per person in order to be at the same level as, you know, the other programs we support, and I should say this is a gross oversimplification, but maybe helpful to just conceptualize what's going on. But if it, instead, is not 10 times as impactful per person, but only five times as impactful per person as other programs, then it's literally like half as much impact per dollar as other programs that we could support.
And given that we have limited resources, we want to direct it to those programs that are having, you know, the highest level impact and we wouldn't want to direct money where it could have half as much impact. And so that meant that these questions around, what's the underlying mortality rate, what effect will ALIMA have, what effect comes from treating these conditions was really a critical [00:18:00] question. Because, you know, the difference between, going back to my example, five times as much impact per person as the alternative, or 10 times as much impact per person as the alternative, in many ways, has a significant bearing, if not all the bearing, it's the crux of whether or not this was a grant that we should have made.
Rosie Bettle: Yeah, I think so, exactly. You're reaching fewer people maybe per dollar because it's a bit more expensive. But you know, a high percentage of them are actually ill and you're directly targeting them. Then there's this fuzzy, difficult-to-answer question about whether there are longer-term impacts through things like strengthening the healthcare system, improving care seeking in that local population, these kind of more qualitative aspects.
Elie Hassenfeld: I want to track back and talk a little bit about just the process with this grant. Rosie, at the very beginning you mentioned that now you're working on malaria, and at the time that this grant was being investigated, you were working on, you know, rapid response team, which was basically our attempt to dedicate some people [00:19:00] to broadly looking at what had changed due to aid cuts and ensuring that we could move money where we found needs, even if it didn't fit neatly into our preexisting organizational structure.
So I'm curious, how has this investigation of this particular grant for ALIMA in Cameroon similar and different from other grant investigations that we did? We mentioned some of these at the beginning that we hit go on day one as opposed to having all this pre-work that led to the investigation plan. But you know, what are the other ways, if any, in which this as a quote rapid response grant was different?
Rosie Bettle: Sure. So things that were different, I think because we were aware we're on a tight timeline and this was gonna be a program that would be fairly difficult to model out due to the complexity, we used a strategy of, let's use three simple, back-of-the-envelope calculations, and see if we land in a similar ballpark across those models to try and kind of triangulate in on the impact per dollar. [00:20:00] Versus doing, you know, a full-fledged, thorough, detailed, cost-effectiveness analysis that we typically do.
So, for example, one of our simple models was kind of going from top down, if you like. So it's saying, if a population similar to the far north Cameroon has access to ALIMA-level healthcare versus one that doesn't, what is the overall mortality effect to that population served? And then one was kind of more bottom up and was saying, if we predict the number of people ALIMA's going to treat for all these different conditions that we think are really driving the mortality effect, you know, malaria, malnutrition, how many of those people are they going to treat? What's the untreated mortality of those people? What is the efficacy of the treatment? What percentage of those people would get treatment even without ALIMA? So you can then tot up, you know, across condition to get a sense of lives saved per dollar. [00:21:00]
So, yeah, one theme was using a few different simple models to triangulate in, versus one more like in-depth model approach. So that felt a bit different from the modeling side. Otherwise, I would say we took an approach of try to talk to quite a few people fairly early just because we knew about the time sensitivity.
We spent a lot of time talking to ALIMA, trying to make sense of what was happening on the ground and get, you know, that mental picture in your head. We spoke to an external expert who also runs an NGO that focuses on improving treatment access in very underserved populations.
We spoke with a hospital director in Mokolo who's not employed by ALIMA, but who could speak to what the program looks like on the ground.
Elie Hassenfeld: So you were kind of comparing our normal process to this one, where in this one we used the approach of triangulating based on a few simple models rather than the big in-depth model. And I'm [00:22:00] curious, like going forward, should we do that more? Do you think that this teaches us something about a way to move more money out the door, like more expediently, more quickly to help people.
Again, I want to know what you think about this, but maybe the, you know, three simpler models triangulating are getting us, you know, as good an answer or maybe better than the giant model that's like fully in depth and sort of the canonical answer to the question. And, you know, should we take that away for other work we do in the future. Or maybe we shouldn't, and if not, why not?
Rosie Bettle: I think that's such a good question, and I feel I'm speaking for myself, like other people at GiveWell probably have different opinions, and I feel like this is something we're grappling with.
I do like the approach of triangulating with a few different models, because I think that in forcing yourself to think about a program from a different perspective, different angle, different layer, you sort of, at least for me, I often realize that I've been thinking about something a little bit wrong or I've been rabbit-holing into one particular parameter that isn't actually so key to what the [00:23:00] program is doing.
At the same time, though, having a full detailed cost-effectiveness analysis, that's really our model of what exactly is happening, you know, and trying to put in every single thing that we think is majorly affecting a program's impact.
So, I don't know, maybe I would lean to quick models triangulation when you need to make a decision pretty quick, full model when you are really trying to understand how a program works. Maybe with a quick triangulation, with a simple model too, because I personally love simple models.
Alice Redfern: And I think part of what made it tricky with this grant was that we had this triangulation, but all of the answers that we were getting were kind of pretty close to our decision thresholds.
So triangulation is extremely helpful when it gives you multiple answers that convince you that something is really, really cost-effective. What we were getting was multiple answers that were telling us that this was a very good program that was very close [00:24:00] to or near our bar for being extremely cost-effective.
And I think that was what kind of made some of these questions around mortality and specific inputs become more and more important, because we felt that real uncertainty of where this program actually lied. But in situations where you have something that is looking really good, I think that triangulating a couple of simple models is a great way to move faster.
Elie Hassenfeld: And so then what ultimately got you to yes on this grant. You know, how did you make the decision, reach the conclusion, given that it was close, to move ahead?
Alice Redfern: I think it was a few things. I think it was, like we said, everything was ending up in the same place. And they did talk about this in the Planet Money podcast, that there's some weight in the fact that three completely different approaches end up in the same place. And that place is saying that this is very good. I think that was a big part of it. We didn't have one model that was saying that this was, you know, considerably worse than the others, they were all similar.
[00:25:00] And then the other part was just the more that we thought about the other things that weren't in the model, and I remember Rosie and I sitting opposite each other in London, and being like, but what about this and what about this? And like, is it in there? Are we double counting it somewhere? Like maybe it's not. And you know, I think that the cumulative list of things that we felt like we were just not capturing made us feel good about the estimate being maybe on the lower end of the kind of very cost-effective programs that we fund normally.
Elie Hassenfeld: It was like triangulation gets you to pretty good, but not a definite. And then looking at the, whatever you want to call it, like qualitative factors or unmodeled factors, that was enough to push you above. And so then you get to the answer of yes.
Alice Redfern: Exactly that, and this was spoken about in the podcast as well, speaking to people on the ground and just, you know, it does help to hear from a third party about their perspective on what will happen without the [00:26:00] program. And, you know, we're creating this version of what happens about the program in the model. To be able to hear someone talk through what that means to them in real life and put that to the numbers, helps you feel better about it as well.
I think, yeah, I have it here, there's about 80 staff that were expected to be lost from a single hospital, and that's a real mixture of support, these contracted staff, ALIMA directly hired staff, people at all different levels. And I think that that can make it difficult for us to evaluate as well, because it's not a clean picture.
I remember us having these back and forth calls with ALIMA trying to go into the details of who are all these different people, what are they doing? And it, again, it just doesn't fit cleanly into a single model of a program, which is how we try and think about things.
Elie Hassenfeld: Right. And it really illustrates the messiness and the challenge of just trying to figure out what it is that you're getting when you donate money to an organization running this kind of program.
And I think for a long time, the decision that we made at GiveWell, I made at [00:27:00] GiveWell was, you know, this is years and years ago, but to say like, we just don't have the capacity to deal with something like this because we know it'll be too hard. And I think something that is, I don't know really great from my perspective is that with this additional capacity, we're in a position when the moment really needed it, to be able to look at something like this and say, it's really challenging, we're not gonna be able to do it in the clean, quantitative way in the way we ideally would, but obviously we can get somewhere and then we're able to direct money where we think it will do a lot of good.
So, I'm glad that we were in a position to be able to do that earlier this year, I should say differently, that you were here to be able to do that earlier this year.
Maybe just as closing, I'd love to hear what's happening with this grant now. This grant was made several months ago at this point, and so to the extent that you can share what's happening now, that would be really great.
Alice Redfern: Yeah, sure. So, it's actually a couple of months now since we last caught up with ALIMA, we're catching up with them next week, coincidentally. But as of when we last spoke, everything was up and running again. There was a lull as we were doing the investigation and then making the grant. And then after that it [00:28:00] took them a month or two to kind of renegotiate reentry with the Ministry of Health, make sure everything was reinstated again, and reach back up to the scale they were operating at before.
They've also started operating a few more mobile clinics, in a few places rather than operating directly in the facilities, they've been able to do more outreach than they were doing before. So they've kind of adapted the program as they've reopened it. And they are really, as far as we've heard from them, on track with treating the number of children that they were hoping to treat through this grant, which is good news. There's not been any major curve balls, so far.
We're receiving the data from them in the next few days on how the first six months have gone, so I'll have more information soon.
Elie Hassenfeld: Anything else that we didn't talk about that, Rosie, Alice, either of you want to add just to share about this grant before we sign off?
Rosie Bettle: Something that hasn't really come up that I found really interesting was, especially in the context of thinking about counterfactual coverage, what happens absent ALIMA, [00:29:00] is that we kept hearing from people that with ALIMA, there's a big uptick in care-seeking behavior. So people want to come into the outpatient clinic to get the ready-to-use therapeutic food for their kids, for example. They'll come there if they know that ALIMA's there. And while they're there, they also, you know, get their catch-up vaccinations, maybe basic health check, this kind of stuff.
And it was kind of interesting for me to think of it as both a push and a pull mechanism, if that makes sense, for people to be like, okay, ALIMA's there, they're operating, I'm going to go to the hospital to check my kid, et cetera, et cetera.
Elie Hassenfeld: And I think that's just another reason that, you know, from my perspective, when programs we look at seem close to our funding bar, I think it's appropriate to assume that we're being slightly conservative in our estimates, and it can just maybe give us a little push to move some money to a good organization, you know, more quickly.
I say that just because that's one of the reflections I have thinking about this conversation. I mean, I look back [00:30:00] at this, and I do think about six weeks as, from my perspective, extremely short to go from nothing to grant for $2 million. And so it seems like we're operating on a very fast timeline in general and doing a lot.
And I hope it also gives us some confidence to be in a position to just keep moving money out the door when we see the programs that are going to be really great. Thank you both for talking this through with me, this was really great, so thanks for having this conversation.
--
Elie Hassenfeld: This is Elie again. Thanks so much for joining us for this conversation. I think this grant investigation and this conversation really illustrate a few things that were particularly salient in our work in 2025.
One thing we faced was the challenge of responding to unexpected events and changing our approach to be able to ensure that we could direct money where it could do a lot of good. Before the cuts happened, programs like the one ALIMA was supporting in Cameroon were funded. And because of that, it wasn't the kind of program, this nutrition plus primary healthcare [00:31:00] services, that we had looked at extensively. But we knew that there was a significant need. And so when we saw that, we created the rapid response team and jumped in to be able to assess opportunities like this that we wouldn't have looked at in the past.
I think it also just illustrates how much GiveWell has evolved in the last few years. It's the case that several years ago, we would've looked at something like this opportunity and just said, it's too complicated. We don't have the capacity, we don't have the experience, we don't have the capability to make a good decision here. And so we need to focus on a narrower set of programs.
And you know, today we've grown the team a lot, and so we're in a position to look at programs like this, and that just means that we can do a better job directing money where it will help people in need around the world.
I think this conversation also shows just how challenging it can be to quantify the impacts of a program. GiveWell likes to say, you know, this is the cost per life saved of a program. And as best we can, we try to communicate that we have significant uncertainty about those numbers, even [00:32:00] in programs that are relatively easy to model. A program like this is just so, so hard to come to a reasonably confident estimate about the impact per dollar spent.
And then finally, I think one of the things that we did in this grant was operate differently. You know, we don't have a full in-depth, detailed, canonical model that is a reflection of every step in the causal chain from donation to impact that we've worked through over many years. Instead, we had to do something different, and instead used a few different approaches to triangulating the benefits or some subset of the benefits that convinced us that this was a place to put money that would do a lot of good.
As we continue to move forward in a time with larger, urgent, and significant needs, I know that we're going to keep thinking about ways that we can take money from donors and get it to organizations that are going to help people in need as effectively and expediently as possible.
As always, we [00:33:00] appreciate your financial support and if you want to support us, there's a few different ways you can do that. Our top recommendation for most people is to give to our Top Charities Fund. These funds go to one of four organizations that we have high confidence in their impact, their track record, and their ability to deliver high impact per dollar.
If you want to support the full scope of our grantmaking, you can support the All Grants Fund. These grants can carry higher risk, they're often organizations we know less about, and programs that we're getting to know, like this one with ALIMA. On the other hand, it gives us the flexibility to support programs that can be incredibly impactful and allows us to go beyond the scope of just this small set of organizations that have made our Top Charities list.
And finally, for donors who have known us and trust us, you can give unrestricted, and that potentially supports GiveWell's operations. We've been fortunate that for the last few years, the amount of funding that we've been able to raise for our operations has outstripped our needs. And so we've been able to move those [00:34:00] funds over into our grantmaking pots so we can direct them to places that need them. That operational support gives us the confidence to keep growing the team and ensuring that we can deliver great research and great impact for people around the world.
As always, thank you so much, and if you have any questions, please email info@givewell.org.