How do you know how to save a drowning child across the world?

Compassion, community, and cost-effectiveness

I’m sympathetic to thoughtful criticisms of effective altruism — the movement that argues we should focus our donations on cost-effective global solutions such as malaria medication — particularly, as is discussed in this recent New Atlantis article by Ari Schulman, the criticism we know the problems and solutions in our own backyards better.

But the often-presented alternative is to focus on local charity in high-income countries, which to me is morally inadequate in our globalized, interconnected world plagued by poverty. I argue here that even if you share certain criticisms of effective altruism and believe strongly in community-focused charity, there’s still a compelling reason to donate to cash transfers or community-based organizations in low- and middle-income countries.

The “drowning child” thought experiment

I can't fully explain effective altruism since I'm not even sure what the term means anymore, but I can explain the drowning child thought experiment, from which comes the basic argument from which much of the movement flows: that you should focus on cost-effective donations in lower-income countries since that's where your money goes the furthest to save lives.

The “drowning child” thought experiment, put forth by the modern utilitarian philosopher Peter Singer, though as this article points out which can be seen to be building on the idea of compassion from the ancient philosopher Mencius, is that if we saw a child drowning nearby we should rescue him, and following that, "Would it make any difference if the child were far away, in another country perhaps, but similarly in danger of death, and equally within your means to save, at no great cost — and absolutely no danger — to yourself?"

Actually, the argument follows, we actually do have the opportunity and obligation to save dying children across the world. Children are dying of malaria every day, and by donating some amount of money you can prevent this. If you agree that the drowning child far away matters as much as the one close by, you should focus your donations on what can save the lives of/help the most people, regardless of where they are. And since there are more acute problems and your dollar goes further in addressing them in lower-income countries, you should focus your donations there.

The organization GiveWell¹ works to try to figure out the most cost-effective things you can donate to — an estimated $5000 donated towards malaria-prevention medicine or infant vaccinations in sub-Saharan Africa can save a life. The charities GiveWell recommends are tested by randomized evaluations (I'll get to this in a post soon) and are highly scalable; they can absorb a lot of donor funds.

But we know best how to help those closest to us

The latest criticism of effective altruism by Ari Schulman in the New Atlantis introduces a version of what I think is one of the more compelling arguments against this in practice, which is, okay, we know how to help the drowning child in front of us. We know the problem; we know the solution. But across the world, we don't know the full problems, or the range of solutions, or the trustworthiness of different actors. We might be skeptical of the estimates of an intermediary. And our time and money is limited. How do we actually know that our donation is helping someone on the other side of the world? It makes sense to focus on something we do know.

In the words of the author:

Given a world where doing good, like doing anything, takes scarce resources of money, action, attention, and information, the pencil lesson can be extended some ways. I could imagine that, between saving children from starvation in China or helping run a food bank for the merely hungry in my own neighborhood, even though the faraway situation is objectively more weighty, I can do more good per unit of effort by volunteering at the food bank. Or I could decide that I should help a family in crisis in my neighborhood, rather than leaving that to a philanthropist in China or a well-meaning aid program in my state, because even though those actors have more funds, I am closer to the situation and have a fuller view of the problem.

This is a compelling argument to an extent, especially to those of us who have seen charities wasting money and are a bit concerned that the money we donate is not actually helping people, especially if we are far from the problem.

I find much to agree with in Schulman's essay. Most relevant here, I am not a totalizing utilitarian; I donate locally as well as globally; I think the world is better for having people care more about their families than strangers, for having sentiment.

But, even if we cannot help them perfectly, I still believe that all drowning children are equally valuable.

Following from this, I think criticisms of effective altruism often end up with a conclusion that is too far in the other direction: the conclusion that we only have moral obligation to people in our immediate circles and thus should focus on parochial charity, a conclusion that does not leave room for moral concern and yes, even obligation,² for the global rich to people living in poverty far from us.

I don't think any argument that focuses solely on helping within communities that we are already in — communities that are, even in the US alone, highly segregated by income; and are globally even more vastly unequal — adequately addresses the moral ill that is global poverty.

Source: Our World in Data: Median income or consumption per day, 2023

I think it is good to volunteer and donate to issues in your relatively wealthy town. But if you only do that, it raises the uncomfortable question of — what of other communities that you will never be a part of, faced with a dearth of funds, communities where everyone is poor and lacks access to medical care and so even if everyone living there had the most neighborly, altruistic impulses — children would still be dying of malaria and tuberculosis?

These communities exist in immeasurable number, even if you can't see them. Their lives are valuable. And I think people in high-income countries with disposable income do have an obligation to them.

When I lived in India and Zambia, other Americans used to ask me whether it bothered me to be so close to so much poverty. And I said of course it did. But at least I could give money to people on the street or buy them water on a hot day. It wasn't changing the world, but it could ensure that someone could eat for a while. Did it not also bother these Americans that these places existed and they lived far away from them and were trying to forget them?

Small circles of moral obligation may have once sufficed. But in the modern world, both our connection to and our ability to reach people across the world have become more salient. Even if we ignore people in other countries, we are indeed connected to them: we benefit from their labor with our mass-produced clothes and shoes and phones; our collective energy consumption contributes to making their summers scorching and harvests unpredictable.

Building trust and sentiment over distance

I think strong communities are very important and that it is usually true that people close to the problems know best how to solve them. As I said, I donate to local organizations in my relatively rich city. But with the vast amounts of poverty and inequality and interconnectedness in the world, I do not think it is enough.

Many of the organizations with highly scalable and measurably cost-effective solutions do one or two things really well — whether it's in health, or agriculture, or education — which they can then scale to a number of different countries at low cost. I think this is great and that's why I've dedicated much of my life to it, but while I don’t share these hesitancies personally, I also understand the arguments from certain effective altruism critics that they are hesitant to donate to these sorts of large, scalable organizations because every individual has very different needs and people in their community will know their needs best, or that they simply don’t agree with or trust the methodology of intermediary charity evaluators. And I definitely agree that not all good things are scalable or easily measurable.

But I implore people who have these sentiments and hesitations about spreadsheets, too, to think about the other communities across the world where people are just as motivated and civically minded as your own but have much less cash and unfortunately, a plethora of problems that high-income countries do not have.

And so is it possible to find some sort of synthesis between donating to community-focused organizations and cost-effective solutions in lower-income countries?

Giving cash is a great step in trying to address this from afar, but it's also true that individual cash transfers can't do everything. An individual probably isn’t best poised to set up functional hospitals or libraries or schools; communities and organizations and governments are usually better than the individual at doing larger projects.

I know that there are a lot of community organizations in low-income countries that are cost-effective, understand what people need, and work to build stronger communities. Local organizations and charities all over the world are providing life-saving medications for children and social services for older people and education support for young adults. They're potentially more in need of marginal funding, though they can absorb fewer funds, than the larger-scale organizations that, excitingly, are already connected to many wealthy donors.

The problem is rich people like myself (and probably you) often don't know about them or know how to find them. And as their impact is fairly direct and based on their knowledge of a community, it wouldn't be a good use of funds to scale them or do a randomized controlled trial on their impact, but it is important to know that they can use the additional funds that they're receiving well — just as I'd want to know before donating to an organization in my own community.

Something like GiveWell's (not up-to-date) guidance for vetting charities can be extremely helpful for known unknowns: if you live in Washington, D.C. or Nairobi and want to vet a local charity — or a charity across the world you know through your family or religious community or friends — these questions can be great to ask. But it doesn't make unknown unknowns known: if you live in D.C., how are you even supposed to know about an organization in Nairobi unless you happen to know someone who works there? That's, as criticism of effective altruism says, hard to know from the other side of the world.

One of my motivating questions for a long time has been to try and figure out how to bridge this gap of knowledge and trust and indeed, sentiment, between people with money who want to give, and the local organizations working with people most in need of money/services, especially across countries.

Let me know if you're also thinking about this. In the meantime, if this bothers you as it bothers me, donating to direct cash transfers is a great option.

1. Note I have donated through GiveWell to their recommended charities

2. I think the "obligation" part is where I may diverge from Schulman, although if he for some reason ever reads this blog he's free to correct me if I'm wrong. I think he would agree with the moral concern part, as he states: "GiveWell’s ratings and rankings and spreadsheets have their place. Every one of those 215,000 lives saved mattered. Their plight does call on us, even from afar. We should celebrate this work, and if more is to come, celebrate it too."