Far-off funders and effective global philanthropy

Originally posted on my Substack

There’s a new wave of philanthropy about to come out of Silicon Valley. Nan Ransohoff, in “The third wave of American philanthropy”, notes that we might soon have an additional $37B of philanthropy per year from Anthropic and the OpenAI Foundation. Some of the soon-to-be-rich have — as have I — been influenced by effective altruism, and so I expect some of this to go to global development.^[1]

What will happen with all this money?

Ransohoff notes that we might need more philanthropies, and fast. The article, which echoes other views I’ve heard, posits the solution to many of the world’s problems as more philanthropies, founded by smart people who can think like tech VCs and deploy money quickly and effectively. The article isn’t only about global development — and I don’t speak for anything but global development — but does call out certain organizations that either completely or partially work in global development as examples of what we need more of.

Overall, we are in a somewhat more positive place than last year, when the shutdown of USAID dried up funding and left many people dead. Quick actions such as cancellation reversals meant that the worst case scenario was averted.

But there is still much to be done — US humanitarian spending, for example, has sharply reduced; the overall cuts and pauses led to an estimated 750,000+ deaths in the first year. I think private philanthropy can do well to fill some of this gap — my old USAID team is, for one, back as a private philanthropy, the DIV Fund — but I don’t share the optimism that more tech VCs are what we need for global development philanthropy. If anything, global development philanthropy needs more founders and grantmakers further from Silicon Valley to maximize impact.

My point is not to criticize one article in particular, but some broader trends in my small corner of the world — evidence-based global development philanthropy and grantmaking, with its close ties to economics and tech.

This corner of the world puts a high premium on the idea of the “smart generalist”— a smart person, usually from and/or based in the US or Europe, who makes grants across countries and fields based on what they can understand with numbers. I was one for two years and I don’t think I was the best person for the job. Here’s what I think we can do better.

***

If I had to summarize this philanthropy sub-field, it’s that we try to fund things that are proven to be effective at improving people’s lives. We also fund research on different programs and policies to try to figure this out, using methods from economics and other social sciences (such as randomized controlled trials). Governments and nonprofits are operating on limited budgets, and the idea is to fund the things that are most effective and cost-effective at improving people’s health, or learning, or income — if we find it works, scale it; if we find it doesn’t, refine or stop doing it. Evidence-based philanthropy appeals to many people in tech and other quantitative fields because everyone wants to donate to something that works, and this provides further quantification of impact — estimates of exactly how many lives a program has saved.

Some of the big, scalable successes have been malaria nets, Vitamin A supplementation, and direct cash transfers.

I’ve worked in this field, defined broadly, essentially my whole career: in government, nonprofit, and media; as a researcher, a writer, a grantmaker, and now, partially an implementer. This has been mostly focused on global development.

I really believe in the core of this — trying to measure a program’s success not just by (still important) intermediate outputs like “we built this many clinics/schools”, but by real life outcomes like longer life, or higher incomes. I think a lot of the organizations founded and funded by US- and Europe-based people in the evidence space have done tremendous good.

Like others in the field, I am really passionate about evidence and I want to make the field better! But over my career, I’ve become convinced that what global development philanthropy needs to improve itself isn’t just better quantification. We need better decisionmakers — ones who aren’t “smart generalists.”

I don’t think evidence-based global philanthropy is always finding or funding the most effective or scalable things, and I posit that this is largely because most people in the sector are “smart generalists” who are still from and live in the US and Europe, which means that networks, universities, and funders are largely far from the countries in the rest of the world where these programs and organizations are.

The problem that faces the US- or Europe-based smart generalist, is that if you are in an ecosystem of other people like you, most of whom are probably very removed from the problems that the organizations you are trying to fund are trying to solve, you’re simply not going to know all the best solutions. We don’t know what we don’t know!

I posit that a very effective way to use some of this incoming capital would be supporting more founders and philanthropists from and/or based in Africa, Asia, and Latin America, and specialists in their fields.

***

The smart generalist often does not know much about the sector, country, or local government context, and has to make quick decisions on limited information. A grantmaker will never have full information, but it’s a lot easier to make a good decision quickly about a community health worker program in India if you’ve worked as or with community health workers in India, instead of having to learn about a whole system and country in a couple days of desk research and interviews. Limited knowledge of context may also lead to uninformed decisions about whether research will actually influence policy and programs. You can fund the most rigorous research, but if you don’t have buy-in it will do nothing in the real world. Working with the people involved in designing the program and making decisions is key to making changes and scaling. This applies to the private sector, too. If you don’t understand the market, it’ll be harder to figure out whether a startup has a good idea or not.

Networks are also important. Small, US- and Europe-based networks make evidence-based grantmaking worse on its own terms. The smart generalist is often not aware of the effective organizations and solutions that might be out there but they don’t know about — the unknown unknowns. It’s harder to evaluate promising organizations if they’re not already plugged into your network.

A lot of effectiveness-oriented grantmaking, in my experience, ends up being aimed at what’s easy to understand for elites in high-income countries, when those are likely not all the most effective or cost-effective organizations (though some of them are!). As a grantmaker, I was always thinking of how many great organizations we were missing because neither of us knew each other. Yes, it’s harder to find effective organizations that might be based in other countries and not within a grantmaker’s professional network, but that just means that grantmakers might need to spend more time to find them (which should increase overall expected value!) and should in many cases be from and/or based in these places to make this easier.^[2]

While admittedly I don’t have an empirical evidence base (maybe some of the philanthropy money should go into studying this!) — one of my readers pointed out the counterargument that e.g. the Gates Foundation has demonstrated impact (though it does have many global offices, and so, like USAID, many of its activities might not necessarily fall under what I’m talking about) and we don’t know the counterfactual of e.g. a more localized version; and increased local funding could open up other problems — these are two reasons I think far-off funding is not finding all of the most effective organizations:

1. Far-off funders fund people who are like them or understandable to them, which can lead to questionable grants being made to ineffective foreign-founded organizations that don’t understand people’s needs or paths to sustainability or scale in a certain place. While I’ve unfortunately seen this happen in practice, it theoretically doesn’t need to happen; I’ve known some rich-country based grantmakers who work hard to try to fund locally grounded organizations.
2. Even far-off funders who try to fund locally will simply not be able to find all the most effective organizations because the knowledge transfer and networks aren’t there; the organizations don’t know about them and they don’t know about the organizations. Unlike (1), this seems to me a fundamental flaw of the far-off grantmaking system that I see as best addressed by more local grantmaking.

In my own experience, one of the joys of moving to domestic work — similar to when I was in Zambia, working on projects in Zambia — is that I am able to go deep into and understand my projects better technically. I, and many people I know and talk to daily, use or will use the services I work on. I have a good sense of how research will or may inform implementation, and I am learning more about my fellow state residents and the workings of the government and policies every day. All this grounding makes me much, much better at my job than I would be otherwise.

The long-term expected value gains in social terms could be tremendous if we could move more of this incoming philanthropic capital into organizations solving existing problems from closer to where they work.^[3]

So what should we do?

I didn’t write this article because I think I have all the answers; I wrote this because I think that there is sometimes too much confidence that we — evidence-based funders, tech employees, philanthropists in rich countries — know the solutions to all the world’s problems just because we are smart, like global poverty can be solved with the right amount of brainpower. So these recommendations are quite uncertain, but focus on making philanthropy a bit closer to people who may be closer to the answers.

I also want to note that a lot of these things are already happening, and a huge props to the philanthropies and other organizations that are doing this. I would just love to see more of these, and I think they could make evidence-based, effectiveness-focused philanthropy truly more effective!

1. Fund philanthropic founders and grantmakers from and/or based in low- and middle-income countries

Funders should support more founders and philanthropists from and/or based^[4] in low- and middle-income countries to allocate funds, and existing funders should hire more locally. Even though it’ll be more work for philanthropists and recruiters up front, it will pay off in the medium- and long-run. Networks are not going to stop being important for grantmaking, so it’s paramount that many grantmakers be from and/or based in the countries that they are working in. In my experience, recruiters in the evidence space continue to recruit from a small pool of candidates, often based in the US and Europe, because this is what they know. And we are fine candidates! But I do not think we are the best.^[5] I worry that this “third wave of philanthropy” will likely tend towards hiring from the same pool of smart generalists — but I hope today’s funders instead surpass what’s happened in the past.

I think funders and recruiters could learn something from civil service hiring tests (and hopefully improve on them) that try to score candidates on somewhat objective dimensions. And the criteria should not just include networks, math tests, or elite universities! Extra points should be given for actually understanding context and how different industries and governments work in different countries, not just being able to do math and sound smart. Back-of-the-envelope calculations are not so rigorous anyway and based on lots of assumptions. Someone with a lot of experience in, e.g. healthcare in Ghana will come in with better assumptions to think about whether a grant will be successful than me, a smart generalist.

I understand generalists are sometimes needed, and I’m not arguing that people like me have no place in philanthropy. But consider that someone with experience in healthcare in Ghana will inherently be something of a generalist already. They will likely be better than me at making funding assumptions both in healthcare anywhere, and any other sector in Ghana. After nearly a year deep in AI product and state government, I am confident I could be a much better grantmaker in either AI or government efficiency than before. Hire me for this, not for finding the most effective healthcare solutions in Ghana.

2. Support regrantors close to the work

Ransohoff brings up the idea of regrantors — “smaller, more specialized, capital allocators” — which I think is a great idea. It’s an idea I’m mostly familiar with from AI but there’s no reason it couldn’t also be used here. If a philanthropist is just an individual, or a philanthropy that has a lot of funds to deploy but for some reason can’t hire lots of people, they could reallocate money to people who are familiar with the space. For example, you could have one regrantor be someone who has worked in the health sector in Uganda for their whole career, and another who has worked in agriculture in Brazil, and they could quickly deploy the money to organizations they know are doing good work and are in need of cash.

3. Fund cost-effective local organizations and share what you learn in your research as a public good

Hopefully 1 and 2 would lead to funding more highly effective local organizations in low- and middle-income countries. Many yet-unknown organizations will have high cost-effectiveness, especially where cost of living is low.

The way I’ve tried to think about this is — it’s not possible to be reaching everyone, but you should be trying to get closer to the people who can solve the problem. I might not ever learn of just-beginning entrepreneurs from small villages in Zambia to fund, but I could, theoretically, fund one of the people I know in Lusaka in the Zambian startup ecosystem as a grantmaker themselves who would then be plugged into a community of other entrepreneurs just starting out. And of course if I lived in Zambia again I could go to different entrepreneurship events myself and therefore become a much better funder.

Following from this, this could include sharing sourcing, expanding the pool of legitimate organizations that are known. I thought this was a great idea from “The legibility trap: how “fund the most rigorous thinkers” can quietly select for the wrong people” by Justin Steele, with whom I had a good discussion in the comments and whose post I’m sure influenced the ideas in this essay in other ways.^[6] (It’s always exciting to encounter another article talking about something you’ve thought and cared about for a long time!)

If grantmakers closer to the work can do the work of sourcing potential grantees and share this information as a public good, then rich-country-based philanthropies (I don’t think these are going away) can have pre-diligence work done for them on harder-to-find organizations and consider them, vs. funding either the small pool of known orgs, or only their friends. Diligence work then compounds; funding can flow faster.

Some of these organizations may not be highly scalable, which makes it harder if the goal is to deploy a lot of capital quickly, because you have to do diligence on and fund many organizations/founders instead of one large one. But having more local grantmakers and philanthropic founders, networks, and knowledge sharing, should make this easier. And as mentioned earlier, having more local grantmakers should also make it easier to find organizations with high potential to scale and research that may actually lead to policy change. Plus, not every funder or philanthropy needs to do the same thing! Maybe in the philanthropy-rich world that Ransohoff talks about, some focus heavily on scale, while some focus on smaller organizations that operate extremely cost-effectively.

4. Consider medium- and long-term effectiveness and cost-effectiveness in global development

My hope is (1) and (2) might also lead to funding more organizations and interventions that are cost-effective when considering the medium- and long-term — even if they don’t look as cost-effective in the short-term — though this could also happen independently.

One specific area I think evidence grantmakers in global development often overlook is long-term benefit, which might mean more seriously funding areas like tertiary education, research and development, and infrastructure that have long-term social and economic gains. I know that these are usually funded (and should be in my opinion) by governments, but I think there are high-impact areas where philanthropy can step in, much like it does in largely government-funded sectors in high income countries. I remember one time talking to a friend from Nigeria who said that everyone wants to fund community health workers, but no one wants to fund medical R&D. I listed some of these suggestions in a post last year and I think they’re still relevant:

• Training and working with policymakers on impact and cost-effectiveness
• Various tertiary education interventions, from sponsorships like the Malengo model, scholarships, or investments into high-quality local tertiary education
• Education interventions in general, including pre-K and K-12. They lose out to global health in the current analyses, but I’m not sure how it would pan out if long-term effects were taken into account
• Mentorship, particularly involving people not already from elite universities or [high-income countries]
• Networking and community-building opportunities, particularly involving people who are not already from elite universities
• Expanding the scope of job boards towards high-impact opportunities not based in the US/UK/[high-income countries]

Investment in human capital is something I think one of my former organizations, IDinsight (a data and analytics organization, not a grantmaker), has done a great job at. While founded in the US, IDinsight has made great effort to both hire locally for its offices around the world, and provide experience, mentorship, and networks to early-career data professionals. I personally greatly benefitted from being part of an organization where people had good advice on graduate school and professional development – and my colleagues from India, Kenya, and the Philippines did as well. Of course, this was a win-win for the organization, since my colleagues also brought their knowledge and networks. Philanthropy could fund more global development-specific organizations that prioritize this, and also invest in things like scholarships or fellowships that would help the networks of people from anywhere.

In addition to these I’ve mentioned, Steele noted three to be tested that I would also be interested in — while his piece was addressed to the Effective Altruism community, I think it also applies to the broader evidence-based global philanthropy world:

1. Treat proximity as a measurable input to expected value, not a tiebreaker after it. […]

2. Fund rigorous evaluations of participatory allocation. […]

3. Back bridgers, and define them operationally so the proposal can be tested rather than admired.

Involving broader communities — beyond professional founders or grantmakers — more directly in philanthropy through participatory grantmaking is something I’m particularly interested in, but I don’t have enough experience in to provide any concrete recommendations. One thing I’ve learned from various forms of engagement is that it can sometimes be tricky to balance trying to get involved and/or representative engagement — I would want this all the time in an ideal world! — and not wanting to waste people’s time when they have lives to live. Sometimes people want to participate involvedly in decisionmaking and sometimes they just want the decisions to be made for them so long as there is room to object. That’s why I think that involving people local to a place and who have relevant skills, and experience who want to be founders and grantmakers is probably a good first step, and more community-based deliberations could happen as a supplement.

***

So, do I think in my ideal world there would still be a role for “smart generalists” or in global development philanthropy based in the US and Europe? Well, in my ideal world, we wouldn’t need philanthropy at all. But in the slightly-less-than ideal world, I think yes. Global inequality is still with us, and it makes sense for some research and fundraising and networking roles, among others, to be based in the US and Europe. There are also many people who are based in the US and Europe due to e.g. university affiliation who do have deep knowledge of sector and context. But overall, I think the balance needs to shift for funding to be to be the best it can be.

1. ^{^}

   I, not alone, have some skepticism around how much of this money will go to good causes. But even if ten percent does — that’s exciting!

2. ^{^}

   This paragraph is adapted from my comment on Justin Steele’s “The legibility trap: how “fund the most rigorous thinkers” can quietly select for the wrong people”. Apparently self-plagiarism can be a thing in some contexts? So consider myself cited.

3. ^{^}

   Related, researchers (and research funders who may fund them due to proximity) who are removed from a country or region may lead to research being lower-quality, irrelevant to program implementation, and in the worst cases, ethically questionable. I wrote a bit about why this is last year and could probably write many posts about it, but I’ll just let you imagine: someone shows up at your house and asks you a bunch of questions for two hours about some program you vaguely remember. The questions show a limited understanding of what your life looks like at all, maybe uses the wrong terms or archaic language for things, maybe asks questions in a confusing way, and then leaves without compensating you for your time. You would probably be annoyed. You might try to answer whatever you think will make it end faster. You might even start trolling them.

   Thankfully, surveys don’t need to be this way, but I’ve been disappointed by how many I’ve seen with elements that make them bad experiences for the respondent. This data is what the evidence of effectiveness discussed earlier rests on. You want it to be good! For this, the research team — along with deep technical rigor — should have a deep understanding of the people they’re surveying, the program they’re looking at, and the place they’re doing research in. This is also true for implementers. Too often evidence-focused funders look only at the technical rigor and nothing else.

^{^}

Note the “from and/or based in”! So this would include someone from Nigeria based in the US and someone from the US based in Nigeria, and even probably someone from the US based in the US who spent 10 years in Nigeria. There is a good discussion to be had who constitutes “local” and how this relates to colonialism, and also within-country diversity, such as ethnic group, religion, caste, region. Ideally a repesentative group of people could become founders and funders in their countries and regions. But I kept the language as broad as possible on purpose because truly the bar is below the floor; most evidence-based grantmakers are both from and based in the US and Europe, with a couple of years at best of experience in another country.

^{^}

If this means fewer recruiters and funders reach out to me (and people like me), I think that’s good. I think ending global poverty is so important, that’s why I keep writing about it, and I could definitely see myself working in it more formally again if the right role arises, but I don’t think I am the right person for most roles. I am very happy in my civil service role and I feel like I am the right person for it.

^{^}

The AI writing style made this essay a little hard to read for me, although I deeply appreciated the ideas and also the author’s disclaimer on LLM use. I will use this to segue back into grantmaking and say that I realize AI makes this all harder. While AI ideally can help open doors — for example, a founder with a great product who speaks English as a second or third language can use AI to help write an application for an English-speaking funder — there is now a massive influx of AI applications for both jobs and grants that make both recruiters’ and grantmakers’ jobs more difficult. You know what I’m going to say! This means networks are even more important and can’t just be based in rich countries if they want to make a difference elsewhere. Both the people distributing money and the people trying to solve the problems should largely be based and/or from the places they’re working in.

Show all footnotes