He would not have been born

Thanks Lyndon1, that describes the situation. Like almost 1 in 10 modern men, my sperm weren’t up to the task - at least not the old fashioned way. So IVF it was, the miracle that now feels commonplace.

We landed on the moon and Woodstock changed music forever. But in 1969 a quieter event transformed the world - the first human embryo was fertilised “In vitro”, outside the body. Then 48 years ago, Louise Joy Brown was the first woman born through IVF. This year about a million babies will be born through IVF.

IVF stands at the frontier of rich-privilege, a norm for the middle class in rich countries. In my home country New Zealand, the government even pays for it! Yet every week I meet women here in Northern Uganda who can’t have kids, and could never afford IVF. Even the human “right” to have a kid can depend on where you are born.

If I was born 25 years ago, or in Uganda, my boy would never have been.

Our Risky Choice

We decided to give birth in Lacor Hospital, the best hospital here in Northern Uganda and where I worked as a doc 15 years ago. If you’re wondering, yes this was risky. While we judged the hospital would be safe for my wife, should something go wrong and our boy needed the neonatal intensive care unit (NICU), he would be less likely to die in New Zealand. By my rough calculations, perhaps three times less likely. Even a good Ugandan hospital is far worse than one in New Zealand, for a few reasons

1. Prematurity: Better hospitals keep more premature babies alive. If born between 24 and 30 weeks, our boy would be twice as likely to survive in New Zealand, compared to where we gave birth.

1. Infections. Basic infection control just isn’t as good compared to a high-income NICU. Parents enter the NICU here en-masse to feed their kids and spread bugs around. Super-bugs aren’t tested for. Hands are less-washed. Cutting-edge antibiotics aren’t available. Our boy would be more likely both to catch an infection and then die from it.

2. Fewer staff and stuff. Fewer nurses and doctors, and less high-tech machines mean that problems will be diagnosed and treated later and more mistakes made. Despite staff’s best effort, kids just can’t be monitored and cared for with the same intensity in Northern Uganda as they can in New Zealand.

Months before birth, a couple of people expressed “you’re mad to put your child at more risk”. This might be true, but the overall risk seemed lowish. We wanted to give birth close to our home of 13 years, in relationship and solidarity with our community. Our privilege meant we had the Lacor Hospital “A team” on the case, with a dedicated gynaecologist and midwife. But despite their best efforts, things didn’t go as planned.

He didn’t die once - a routine miracle

“His head isn’t moving” are not words you want to hear. Our boy’s head made it through the cervix, then he got stuck completely. No maneuver, nor three scary suction attempts could move our poor boy past that liminal space between the safety of the womb and our wild world. My wife suffered, our midwife sweated, fear set in.

In most times and places in human history, this would have been the end. For our baby, or my wife, or both. But we were in the right place at the right time, with a life-saving procedure just a theatre away.

We thank the Roman innovators who first cut babies from dead lifegiving mothers. We thank the farmer Jacob Nufer, who in 1500 used his experience with pigs to be the first recorded to save both his wife and their baby. We thank Fransour Rousset, who proposed enfantement caesarien as a routine life-saving surgery. From the 1800s we thank Morton (ether) and Simpson (chloroform) for safer and less painful surgery. We thank Semmelweis for clean hands and Lister for antiseptics. Max Sanger who first decided to actually stitch the uterus closed. Landsteiner and Fleming for blood transfusions and antibiotics.

We recognise your contributions towards saving over 50 million mothers and babies.2

Even in one of the poorest countries on earth, I didn’t doubt our surgeon’s safe scalpel. Humanity’s stacked innovations have made this miracle almost routine.

Our boy was safely born.

He didn’t die twice - Neonatal care

After the exhilaration of holding our boy for the first time, things weren’t as they should be. He didn’t feed properly, became lethargic and turned a bit yellow. A doctor’s assessment was enough to confirm our fears. He was jaundiced and had a fever - he might have the dreaded neonatal sepsis.

We were fortunate to occupy one of four private rooms which had running water, wall plugs and beds - luxuries the 20 other NICU families didn’t have. Every 2 hours as we moved to the NICU to feed him, we tiptoed over carers who lined the corridor. I was overwhelmed with guilt from two angles. First that we paid $30 a day for a private room while others slept on the floor. And guilt at our decision not to give birth in New Zealand, where our boy would be more likely to live. Guilt at our immense privilege, and guilt that we hadn’t used it to the max.

Over the next 10 days yet another wonderful set of innovations helped him pull through. Our resource poor NICU was still equipped with the most important life-saving interventions, and kept our boy warm, well monitored and infection free.

In 1880, Dr. Tarnier noticed that Paris zoo’s warming chambers kept young birds alive, and wondered why we didn’t do the same for sick human babies. Within a few years his colleague Dr. Bodin proved that incubators save lives, but as we’ve seen over the last 20 years with vaccine hesitancy, science alone isn’t always enough. Incubators needed to go viral. So they signed up an “influencer” to make incubators mainstream. Dr. Couney’s business model would be unthinkable today. Mothers got free incubator care, with costs paid for by thousands of tourists who paid entrance fees to see odd-looking premature babies in the strange space-age incubator machines.

Couney (middle) at his exhibit for the Alaska–Yukon–Pacific Exposition, 1909 (from wikipedia)

“Couney’s Child Hatchery” was a wild commercial and medical success and helped catapult incubators into the mainstream, while warming babies become a norm around the world. Progress sometimes needs all of great ideas, science and PR to scale.

But on the newborn care front, we haven’t “won” the battle to the same extent as with the C-Section. 5000 children under one month old still die every day and 500 of those from neonatal sepsis.

Our boy had lethargy, poor feeding and fever, but hospitals can’t just put all lethargic babies in NICU on all out treatment - especially not in low-income settings. Neonatal sepsis has a diagnosis problem. No single test, or range of clinical features tells us if there’s sepsis. A Center for Global Development team believes this test could be a great life-saving bet, and have proposed a fund of 60 million dollars for any company that develops a working test. For the moment this remains a dream, good tests are far harder to invent than vaccines and treatments.

We would have LOVED one simple test to tell us if our boy was at risk of death, or just struggling a bit to get going.

I am Grateful

At a different time, or even right now in a different place our boy would never have been born and might have died twice. We thank the hundreds of innovators and medical staff who took their life saving ideas and made them real. We thank the doctors and nurses who worked with limited resources to help our boy survive and thrive. We thank my mum who struggled with us in those early weeks.

And more than ever I’m grateful for my work with OneDay Health. A friend thought “After you have a kid, your work will feel less important”. But no, it feels more urgent and more important than ever. Even after the worst two weeks of my life, I can’t walk in the shoes of my friends here. I can’t imagine not having two dollars to treat my newborn’s malaria. Or 20 dollars to reach hospital to give birth. Or even not having $5,000 to cure my child’s heart defect.

But I can imagine doing something today to help these babies live, and there’s never been a better day than today.

“If you had to choose a moment in history to be born, and you did not know ahead of time who you would be — you didn’t know whether you were going to be born into a wealthy family or a poor family, what country you’d be born in, whether you were going to be a man or a woman — if you had to choose blindly what moment you’d want to be born, you’d choose now.” - Barrack Obama

Now that we’ve said RIP to USAID, there remain two big piles of non-philanthropic money, which can take the best social solutions to the moon. The first is paying customers. The second is today’s topic, the government.

Scale-through-government has become a panacea for ambitious NGOs. The last slide of pitch decks shows how governments will scale their dream to millions. Smiling photos with ministry officials litter Linkedin.

But I have concerns

Three Concerns about Scale through Government

Government in an ideal world should be the major player who pays for and rolls out the highest impact interventions, not us do-good NGOs. Their core job is to do the best they can for their people. So while economic growth moves a whole country forward, governments should double down on the most cost-effective healthcare, education and income generating activities which improve the lives of thei most vulnerable. We at Oneday Health would love to help governments get good healthcare to everyone, yet we want to embark on this journey eyes wide open. Scale through government It is not the only way, can be morally dodgy and is really hard.

Not the only way

The exceptions might prove the rule, but a number of awesome interventions have scalde huuuuuge through donations, without government cash. International Aid agencies and huge foundations such as Gates and GiveWell still have enough money to “forever fund” important interventions. And we may soon get a new wave of philanthropic go-go-go with Anthropic’s IPO-O-O.

1. Vaccines: The Global Vaccine Alliance (GAVI) scaled up vaccination with enormous success, with 50 million kids vaccinated every year. If GAVI were an NGO it’s 2 billion budget might might be the biggest in the world. Gavi covers 80-90% of vaccine costs in the poorest countries. As countries move to middle income status however, GAVI does force transition to government as payer.
   

2. Mosquito nets: Mosquito net distribution is one of the most cost-effective ways to save lives in the world from deadly malaria. Between 2005 and 2020, Big Aid and Philanthropy scaled up malaria net distribution from close to zero to over 300 million nets delivered for free each year. The Global fund, President’s Malaria initiative and Against Malaria Foundation now spend One Billion dollars a year distributing mosquito nets. Little government money has been spent in this successful scale-up
   

1. Cash Transfers: GiveDirectly spearhead the global cash-transfer movement, which argues that cash transfers help people more than most development interventions. Through a mixture of philanthropy and government AID they 10x-ed their giving in just 5 years, and have given away over 100 million dollars a year since. This philanthropy driven approach has also influenced big aid policy. The UN refugee agency has adopted the “‘why-not cash” approach, which gives cash transfers priority over in-kind assistance.” (https://indevelopmentmag.com/money-for-nothing-the-roles-of-evidence-in-givedirectlys-journey-to-1-billion-delivered.
   
   

Despite this wild success and the mountain of evidence which supports cash, there are very few low income countries that give out their own cash at any kind of scale. The Ugandan government recently gave ten millions of dollars of loans tied to buying livestock, an inefficient process mired in corruption. Free cash would have been way better.

Should governments pay for the best interventions?

Yes, governments should priortise that which helps their people the most. I wonder how much this continuous flow of philanthropy mega-cash reduces government’ incentive to pay. It seems strange that low-income governments don’t pay for vaccines and nets, the cheapest way to save lives. Meanwhile Uganda and Tanzania spend their meagre cash on 300 million dollar specialist hospitals which will serve the rich. Not to mention the 50 million Uganda gave to just 500 MPs this year as a car allowance…

Morally Dodgy

To build relationships with governments, you need to spend lots of money and effort on “sales” to bring them on-side. As Kevin Starr puts it “you stop building more factories, and you start building something that looks like a sales and customer service team.” Through this sales or lobbying process, NGOs will engage in a number of practices that could be considered corrupt in high income countries. I consider them “morally dodgy” at best. NGOs will often…

1. Hand rich government officials clean money through large “per diems” to entice them to attend meetings.

2. Pay for flights and hotel rooms to help officials attend conferences.

3. Pay for media opportunities for politicians to promote themselves while they champion your cause.

Some might consider this a form of “soft corruption”, while these practices also make it harder for other interventions to scale through government in future. You need a sizable “sales war chest” to even be taken seriously by Governments. I might be willing to suck up my self-righteousness and dish up money to govt. officials to help the poorest people get good healthcare, but this feels at best icky and at worst a bit corrupt.

Its really hard

Despite all this buzz about scale through government, there are few examples in low income countries where NGOs have helped government both do the thing and pay for it. In the few success stories, the money and time needed to get there is mind boggling. I deeply respect the Community Health Worker (CHW) movement who helped governments adopt and pay for salaried and supervised CHWs in many Sub-Saharan African countries. But this took a mountain of work - a multi-faceted campaign over 20+ years.

1. High profile experts championed their cause (Paul Farmer, Raj Panjabi)
2. Produced mountains of research to show that “Community Health workers Work”
3. Lobbied the WHO to change their official policy on paying CHWs
4. Formed Advocacy/Lobby coalition CHIC to co-ordinate and lead their charge
5. Spent Tens of millions in sales/lobbying governments
6. Guaranteed hundreds of millions in co-funding governments for a transition period.

For small orgs with big ideas like us, this kind of campaign seems impossible. Fortunately there are smaller govt-as-payer success stories such as Healthy Learners and Food for Educatoin But govt. Still only pays a small proportion of their interventions’ cost, most remains from donors for now.

And one bonus concern if we do succeed…

Poor Taxpayers not Excess Western Money now pay for the impact

We don’t get exponential impact for free, much of the burden shifts from philanthropy to fairly poor taxpayers. In low income countries, the “middle class” along with government debt pays for the impact. Not rich politicians and business people, but wage workers like our nurses who earn $150 a month. The scale graph could be misleading to some as it only displays the philanthropy money spent, while the total spend still tracks the impact line. The advantage of scale through government lies in unlocking exponential scale and long term sustainability. Impact is not magically decoupled from money.

So we better make sure our interventions are really worth the government paying for.

All this said, I’m convinced that scale through government can be be the best path for many great ideas. If you work in government and are up for a chat about getting good healthcare to remote places, please do slip into my DMs!

Every Month

Periods are a challenge to manage anywhere in the world, but in low income rural areas, women must overcome huge physical, social and financial barriers. Every month. Most women in this tricky situation manage their periods in one of two ways.1

1. Reused cloths/rags, not made for purpose
2. Pads - Disposable or reusable

These aren’t great options, and in low income countries women are split between them about 50/502. The worst of the two is reusable cloths which are unsanitary, reduce productivity and contribute to a monthly cycle of worry. On the other hand disposable pads cost between $20 and $50 dollars a year3, too expensive for those living on dollars a day. Reusable pads are cheaper, but carry similar hygiene problems of cloths.

And the scale of period poverty is massive. My best guess is that between 500 million and 1 billion women don’t buy any period products at all.

The Period Poverty Burden

There are huge financial, education and healthcare gaps between high and low income countries. Period poverty is almost a caricature of these disparities. With such a wide range of challenges that contribute to period poverty, it can be hard to isolate which is even most important (we’ll get to that). These include…

1. Financial Burden of buying period products4

2. Health problems caused by unsanitary products

3. Educational time lost

4. Struggle to work and do household tasks during periods

5. Anxiety and reduced wellbeing

6. Environmental cost of disposing pads5

The Menstrual cup

Might solve these challenges6. They’re a one-shot buy which can last over 10 years. Hard physical labour becomes easier, and cups might prevent health problems. Are they just better in every way? These features plus the environmental benefits have led to gradual uptake in high-income countries. About 10-20% of women in high income countries use cups, yet most people in low-income countries have never heard of them. Even among the middle classes cups are but a whisper. 15 of our OneDay Health nurses are trying Asan Cups at the moment - only one had ever heard of them before.

Women in Janakipuram, Uttar Pradesh receive Asan cups, in partnership with Pad & Prejudice

Cups as a “Leapfrog” Tech?

In low-income countries, some technology has been passed over and “leapfrogged” in favour of newer, better options. In East Africa cellphones leapfrogged landlines, saving billions on infrastructure. Mobile money leapfrogged banks. The menstrual cup could leapfrog older “worse” period products. If cups really are much better, and there is a concerted drive to promote and distribute cups, we may have the opportunity to skip straight from rags to cups.

Asan Cups

Asan wants to transform the period poverty landscape for women in low income countries. Their founder Ira was inspired after seeing a domestic worker in India miss work due to period poverty, then designed the Asan cup as a Harvard grad student. Asan sells their high quality cups in high income countries for about $30, then donates cups or sells them at low cost in low income settings. Over the last 5 years they have distributed 100,000 cups through non-profit partners and factories, most in rural India. Through this process they’ve learned a lot about what drives successful cup uptake. Here are 3 key insights.

1) Cup quality is key. If a cup leaks or causes discomfort then women will ditch it, even if the cup was cheap. Adoption depends on a well engineered product that’s easy to use.

2) Peer networks are powerful. When trusted friends and community members use cups, word of mouth can spur a snowball effect. Social proof matters more than marketing.

3) Schools might seem like a good place to distribute cups, but have far lower uptake than community distribution. The special relationship between mothers and daughters can make all the difference.

Challenges

Mass cup distribution and uptake isn’t easy. A number of moral objections and practical challenges stand in the way.

1. Are we removing womens’ choice? Womens’ rights campaigners can balk at a one-size-fits-all menstrual solution because of the understandable stance that all women should have the right to choose whatever method they prefer. Powerful foreigners deciding what’s best for poor women can also seem paternalistic or even neocolonial.

2. Behaviour change: Is always tough. A seachange would be needed and cultural barriers dismantled for cup user to proliferate. Here in Uganda there’s a common, understandable aversion to putting any external object inside the vagina. That said, we’re encouraged by our initial nurse trial where almost all nurses who tried cups liked them so much they’re sticking with it.

3. Cleaning the cups: Can be difficult where water is scarce or dirty.

Are Cups Cost-Effective? - A rough estimate

(This gets into the weeds, feel free to skip)

I’m going to do this suuuuuuper loose BOTEC with these assumptions

1. I’ll compare Menstrual Cups to disposable pads
2. I assume context of rural Uganda, the context I know best
3. Financial benefit DALY equivalents will be taken from GiveWell
4. I’ll will use $7 as the cost for buying and distributing one ASAN cup in a rural area7
5. I will assume the average Asan cup lifespan is 7.5 years8.
6. I’ll convert everything as best as possible to DALY equivalents.

1) Money Saved. One cup costs about $7 to buy and distribute and can last over 10 years, which I’ll discount to 7.5. Disposable pads cost about $1.509 a month in Uganda. With pads costing 18$ a year, menstrual cups might save a woman 135$ over 7.5 years. If we (mis)use GiveWell’s estimate that a 2.3x doubling of consumption is equivalent to 1 DALY, and estimate Uganda’s average GDP per capita of $987 a year that’s

135 / (987x2.3) = 0.059 DALY equivalents averted per cup

2) School days - none gained from cups?

To my surprise, large robust RCTs in Nepal10 and Kenya11 showed no increase in school attendance for any period product use. Women are less likely to attend school during their period and menstrual cup advocates who swear black and blue that menstrual cups helps kids stay in school, but alas the best evidence shows no signal.

2) Health benefits - Bacterial Vaginosis (BV)

There is decent evidence that cups reduce discomfort from BV, but no clear evidence of reduction of UTIs or STIs. This analysis doesn’t take into account the increased risk of HIV transmission for women with BV. This protective effect could increase the DALYs prevented by 10x-100x in high HIV prevalence settings.

Cups vs pads/cloth Claude translated a 10-20% reduction in bacterial vaginosis to a DALY improvement of 0.0008 DALYs per cup. Due to low cup cost and a 7 year time horizon this small benefit might still be meaningful at about $10,000 per DALY averted. As irritation and itching are major BV symptoms, this benefit might be better captured in wellbeing improvement.

3. Improved Wellbeing?

“… for the beneficiaries, the single biggest perceived benefit seemed to be immeasurable
… Although the economic and environmental impacts are immense, our focus groups revealed reduction in anxiety and stress relating to periods.” - Asan Cup survey from rural Kanartaka, India.

Improved wellbeing might be largest benefit of menstrual cups, but this hasn’t been studied much. Women report a wide range of wellbeing improvements in qualitative studies and surveys12, but the other hand some studies didn’t show improvement in mental health scores13.

Cups might improve wellbeing through a range of mechanisms which include

1. Enabling more activities (digging, playing, running) during periods
2. Reduced financial stress around monthly financial and practical burdens
3. Overall increased freedom and empowerment

Loose DALY calculation

So in the end, the only large benefit which comes through is from money saved, but this might just be a reflection of the lack of evidence. Here’s the final calculation…

One $7 cup = Dollar DALY (0.059) + Health DALY (+0.0008) + Wellbys?
= 0.0598 DALYs averted per cup distributed
= 1 DALY averted per $120 spent on cups + ???Wellbys

As a sense check, this would mean 1 DALY averted per about 18 cups distributed, which sounds a bit on the high side to me but not-ridiculous. This is a very low quality, early stage BOTEC, and is likely to be an overestimate of cost-effectiveness.

Great potential - Let’s do Wellbeing research

Menstrual cups are a neglected and scalable intervention which have great potential to improve lives. The missing piece is whether the benefit of distributing a cup is large enough to be worth the (low) cost. As benefits cut across financial, physical and psychological realms, cup distribution could be a great candidate for a high quality study which assessed wellbeing benefits. I’m a bit surprised this hasn’t been done already.

With a concerted effort, it might be possible to move towards menstrual cups as a “norm” in low income countries over the next 20 years. They could even be adopted faster in low income countries than high income countries, just like mobile money and solar panels. But at risk of being really boring, we really do need more research….

To finish, a shameless plug for Asan Cup - Ira and her team really care about period poverty for the poorest women. If you’re in the market for a cup, consider buying an Asan. By all accounts their cup is awesome, and you’ll also help women worse off than you get super-cheap cups.

Conflict of interest: Well soon pilot distribution of Asan cups through our remote OneDay Health centers. Asan is a for-profit company so this is on one level free publicity for them.

To state the obvious... none of these ideas will be the best approach all of the time, there’s plenty of grey area and nuance. I start a little playful, then get a little more serious.

1. Ditch the Cars

Close your eyes and picture the first thing that comes into your head when I say “NGO”. It might be………… a shiny white Landcruiser

But owning cars doesn’t usually make economic sense in low income countries. The ‘real’ market makes this clear. Business rarely buy cars, instead they use public transport or motorbikes. When companies do own cars, its more Corolla than Landcruiser as well.

Cars are often more expensive dollar-for-dollar than in richer countries, fuel cost are high and many NGOs hire drivers, all while public transport is dirt cheap. To move 100km in Uganda public transport costs only $5. For off-road work, motorbikes are far cheaper. Efficiency gains from driving are small and rarely justify the expense, and when necessary hiring cars is easy.

Cars also distort community dynamics. My wife works on a complex peace initiative and fights to stop funders visiting in their cars. Years of hard work and trust-building can be undermined when the NGO Land Cruisers show up. All of a sudden the power gap widens and relationships change. Is this really a locally led initiative? Does the will for peace really come from within the community?

NGO vehicles can also set dangerous norms. GiveWell interviewed government health staff in Malawi for a great Podcast series. Healthcare workers complained that they relied on NGO cars to transport samples and medications between health facilities, which stopped after USAID cuts. But these commodities could be transported on motorbikes at almost negligible cost, samples should never have moved in cars. NGOs set an expensive, unrealistic norm which had an easy and cheap alternative.

2. Fund Solutions not Projects

I mean cost-effective, long term solutions. The most effective orgs do one thing well for a long time. Think Against Malaria foundation and New Incentives. Traditional NGOs like Doctors without borders and Amnesty international also prove the value of focus and experience, outperforming generalist BINGOs with their myriad projects. As you solve the same problem for a long time you gain deep knowledge, learn what works best and become more efficient, the charity version of “10,000 hours” of deliberate practice. We need master NGOs, not jacks-of-all-trades. Charity Entrepreneurship leads the way launching orgs that AIM to solve one particular problem. Solutions are also more scalable than projects as testing and iteration builds efficient “scalable models”, that can be replicated across countries or regions.

8 years ago, we signed up 4 of our Health Centers to a “5 year” USAID funded maternity project. Ugandan women bought vouchers for $1, which provided a full package of maternity care. Health facilities would then bill the project for services provided. But the project took 1.5 years of setup and planning, and 1 year of “closing”, so we only provided maternity services for 2.5 years, half the project length! Central project admin sucked up 35% of the project budget, not even counting high admin costs on our provider side. After the project, some maternity centers closed because they couldn’t function without project money.

Short term Projects with a start and a finish are inefficient. You start from scratch so spend a lot planning, plus hiring, tech setup, procurement, launch meetings, early pilots. All before actually helping people. Most staff have little experience in the specific project activity so waste time figuring things out. Then you pay more just to close the project out. All of this wastes money and time.

As well as being inefficient, the “in and out” nature of projects can lead to dangerous duplication problems. During Covid, World Vision funded a temporary “outreach project” where they paid government workers to provide free healthcare one day a week at rural sites. To our dismay, one clinic set up right next to our permanent health center, where people pay a small fee to access care. People gravitated to the free temporary care, our patient numbers dropped and the facility struggled to break even. A temporary project undermined a permanent solution. To World Vision’s credit, after I contacted them they moved the clinic away from our health center.

There’s also little chance to iterate and learn during a project. You do what you’re paid to do, then stop. Yes there’s always a review and commentary on what worked and what didn’t, but its often not worth much because staff without experience are guessing, and we didn’t test properly so don’t really know what the problems were.

I’m concerned that GiveWell and other Effective Altruist funders too often fall into this “Project trap”. Most early GiveWell money was given to orgs that do one thing well, whether experienced orgs like AMF, Hellen Keller or GiveDirectly, or new efficient orgs that Charity Entrepreneurship founded. Now the situation has changed, with more Big International NGOs like CHAI, and PATH almost “commissioned” to perform time-bound projects.

My org OneDay Health partners with a Vision Spring market solution which sells reading glasses for $3 across Uganda. At the same time @Evidence Action and CHAI launched a project which gave away 25,000 reading glasses for free in Uganda in 2025. If these free drops scale up, their project could undermine the world’s leading eye-glass NGO who have a market-based solution, which has distributed millions of reading glasses to date. I hope there was discussion between Evidence Action and VisionSpring before this started, but even if there was I think this duplication has great potential to do harm.

There are some situations where projects might be the best approach, but in general its better to back existing solutions, and fund existing orgs which do one thing well and have been doing it (or plan to) for a long time.

3. Fund Cost Effective

That 20% of the most cost-effective charities might do over 80% of the good in the charity world remains an anathema to many funders. In fields like healthcare, livelihoods and education, every org should have at least a rudimentary cost-effectiveness analysis of their work, that funders should at least consider.

A silver lining post USAID crash is that cost-effectiveness chat is on the rise. A potential donor (unrelated to Effective Altruism) just asked me to fill out a cost-effectiveness spreadsheet, both surprising and awesome. I think Effective Altruists might underrate the influence they’ve had in helping the wider charity world think more cost-effectively.

4. Fund Bimodal - Test and Scale

Most money should go towards two purposes. Test new ideas (small pots), and scale proven solutions (big pots). Unfortunately much donor money still goes to a muddy-middle-ground of large unproven projects. At the end of the project we didn’t learn much, nor do much good.

Our local community Health Worker knocked on my door and explained the new rubbish collection initiative in our village. Apparently we would take our trash to a designated collection point, then pay the council for our efforts. But right now we just burn the trash for free, and no-one in our community sees trash as a big problem. I asked him if he thought this initiative could work...

“Of course not, why would people pay to move their rubbish? But Germain AID (GIZ) is funding the project and they’ve paid me to tell everyone about it.”

A few days later at a party I struck up a conversation with a GIZ worker about said project. The rubbish project would cost around a million dollars, including importing 5 brand new rubbish trucks. I asked her why they didn’t “test the approach first in one area” to see if it would actually work, before spending a million dollars rolling out a huge rubbish collection project? She shrugged and said “well we just won the Development contract so we’re doing it”. Their rubbish collection plan was neither a test, nor was it proven to work somewhere else.

After a (somewhat heated) discussion about the absurdity of it all she basically agreed the rubbish project was unlikely to work, and we went back to eating pizza and talking about the great weather here.

The project failed spectacularly.

Test

Small pots of money can test cost-effectiveness and viability of new ideas on a small scale. If after the test the work seems doable and cost-effective, we can implement on a larger scale. I think we should test more interventions at a smaller scale for less money. Too many decent ideas are left on the research desk table, that we could test with $10,000 - $100,000.

One caveat is that there are many dubious “feesability” studies that test “Can we roll out intervention X in country Y”. Can this pregnancy blood test panel work in Kenya? Can hepatitis B care be integrated into HIV care in Uganda? The answer most of the time is a straightforward “yes”. Systems are often strong enough The more important questions are often “is this cost-effective” and “is there political will”.

One example of testing “feesability” that doesn’t make huge sense to me is the EvidenceAction Glasses project discussed above. On their website they state

But I don’t think this is very difficult. If we can deliver vaccines at scale, we can deliver eye glasses. Vision Spring have already proved that they can deliver glasses, and through a more complex user-pays model.

There are of course exceptions where an intervention might actually be really tricky and we do need to see if we can do it. For example in-line to clean water supplies is complex to roll out at scale, so we should test if we can do it.

Scale

As smaller money pots go to testing, the big bucks should fund stuff at scale we’re pretty sure is cost-effective and doable. Give cash, give Vaccines, pay people to get vaccines, get teachers better plans, teach at the right level, treat HIV, Prevent malaria, cure malaria etc. Orgs like GAVI, GiveWell and the Gates Foundation generally do this well much of the time. Scale can happen through different, although I agree with Mulago Foundation that the Government and the Market are the most common pathways to huge scale.

5. Lower Salaries

Yes this sounds terrible, seems unfair and pains me to say. But its important.

In most cases, NGO workers are paid 2x-4x more than similar private sector roles. In orgs like the UN and USAID, it can be 10x. The benefits of paying more are immediately gratifying – better workers, high staff retention and less rich-western guilt. But unfortunately harms outweigh those benefits. Fortunately great workers are usually happy with a little more than the market rate, and by paying a bit less orgs can avoid a wide range of potential harms.

a. Less cost-effective work: Often salaries are the biggest org expense, so impact our bottom line, making every dollar do more good for those we serve. If we think our work “needs” a lot of highly paid workers, maybe the work is not simple enough to be cost-effective at scale.

b. Brain Misallocation “In Ugandan villages where non-governmental organisations (NGOs) hired away the existing government health worker, infant mortality went up.” High NGO salaries pull workers away from other important work in government, business or politics. We might think our work is the most-important, but if NGOs suck the top-talent from business and government, they can undermine the very systems we build our cost-effective solutions on.

c. Dissatisfaction at other orgs: While most NGOs have inflated salaries, other NGOs struggle to buck the trend and pay market rates. Workers naturally compare their salaries to other NGO workers and look to jump ship at a moments notice. Even when people don’t jump ship, its not very nice working for half as much money as your mate doing a similar job.

d. Market distortion. “this back of the envelope calculation that we were 75% of the local economy… we destroyed and distorted the local economy completely; as development practice that would’ve been utterly and completely unethical.” Ivan Gayton from MSF.

High NGO salaries drive inflation, from rent to food. This makes it harder for others not in the NGO-economy to survive. Gulu where I live is the most expensive city in Uganda, and its hard to imagine any good reason other than a heavy NGO presence. This makes doing business and attracting investment hard, Gulu has less manufacturing than other similar sized cities. I’ve seen people try to do business here, but quickly move elsewhere as salary expectations are too high.

The education market also gets distorted. The flood of post-war NGOs triggered a flood of “development studies” degrees as young people chased the NGO gold rush. Within a few years most NGOs were gone and the degrees no longer glittered.

With less money in the ecosystem, what better time to recalibrate and pay workers something closer to the market rate.

This is far from an exhaustive list, but what seemed important and “changeable” to me at the time. I’m keen to hear your thoughts, especially disagreements of course!