Cross-posted from Peter McIntyre’s blog.

Why should you care about donating to the best charity?

It’s simple: imagine it is your friends, family, or even you who is in need of help. You don’t want to choose a charity that can tell you a good story about how much it’s helping. You don’t want to choose a charity that helps you a little bit. You don’t even want to choose a charity that helps a lot. You want to help as much as possible. 

Choosing a more effective charity could end up being as good as donating 1600 times the amount. See image 1 about treatment for HIV related diseases and the disability adjusted life years (1 healthy year of life).

Cost-effectiveness of treatments for HIV.

Image 1: DALYs per dollar for treating different HIV related disease

Reason #2: The right donation can save lives; and the wrong donation can accomplish nothing.

Enter effective altruism. Effective altruism is a movement that tries to answer this question:

Of all the ways of helping others in the world, how can I help as much as possible?

And then DO IT.

If this is your first contact with the concept of effective altruism, don’t let it be. I highly recommend one of the following:

  1. Peter Singer (TED) – The why and how of effective altruism
  2. Ben Kuhn -Effective altruism reading material for busy people
  3. Efficient charity: do unto others…

But you know why you’re here.

You already know that you want to help more rather than less. How do you actually do that? Below are 2 suggested approaches, using heuristics (rules of thumb), and looking at the research that has already been done on this topic.

Charity selection heuristicsYoure gonna have a bad time

  • Be swayed to where you can do the most good. My grandfather died of prostate cancer when I was young. The year I graduated from high school, I donated my Christmas money to a cancer charity because I was sad about his passing. A donation wouldn’t bring him back, but maybe it would save someone else’s life. I don’t think that’s the most
  • If your selection is only based on overheads or administrative fees, you’re gonna have a bad time. Imagine a charity where 100% of your funds went to ‘Daffodils for Dogs’?
  • Think about the effectiveness of interventions (and the evidence for that), and then find charities that do these good interventions. For example, there are countless examples of things that sound like a really good idea, but do no good or sometimes even do harm (like the PlayPump). See also: list 1; list 2 . Once you’ve found some promising areas, you can find good charities working in these effective interventions.
  • Thinking about tractability. You need to actually have some chance of succeeding for it to be worthwhile.
  • What will your extra donations be used for? If the Peter Fund does a certain amount of good with $1 million, what would an extra million dollars achieve.
  • Room for more funding. Can they scale efficiently?
  • Neglected. With fewer people working in an area, it’s more likely that it was your donation that helped.
  • Overseas. Charities that do work in the developing world do much more good for what you can donate there. The reasons for this are mainly because governments in the developing world already often guarantee a basic minimum of income and healthcare (more detail here). The Australian government extends someone’s life by a year of healthy, quality life (known as a QALY) for up to $20 000 – $50 000 each; the Against Malaria Foundation for around $100.
  • Transparency. Do you know what is going on with the charity, behind the thin veneer of what they selectively publish to look good?
  • Evaluation. You want to know that a charity rigorously does impact assessments and tacks when one isn’t working.

Promising causes

Global poverty

There is a lot to choosing a good charity. Luckily, there are a bunch of people working on it and we can use their research. They now have 20+ full time employees, write up extremely detailed analysis of charities, and last year people gave $27 million to the charities GiveWell recommended, because they recommended it. At the moment, they advise giving to the following charities:

GiveWell's 2014 recommendations

GiveWell’s 2014 recommendations

Against Malaria Foundation

Distributes long lasting insecticide treated bed nets (LLINs) to stop people getting infected with malaria. On its own, malaria is pretty horrible – it causes someone to have 2 weeks of something a bit worse than the flu. They get a headache, sore muscles, diarrhoea, a fever and generally feel really sick. At a cost of about $5 per net, including all operational costs.

The best bit? For about $3500 you can save someone’s life. Best $3500 you ever spent, right? We also know that bed nets work: bed net distribution is one of the best studied areas of  development studies. The Against Malaria Foundation is tax deductible in Australia, you can donate via their website. Read the full report by GiveWell.

Unfortunately, the rest of the charities listed aren’t tax deductible in Australia.


GiveDirectly gives unconditional cash transfers to the poorest households they can find, mainly in rural Kenya. They do this through mobile phone payments (I was surprised too!). 87% of donated money goes straight to participants. They did a randomised controlled trial (RCT), on their work and found substantial increases in short-term consumption, especially food, and little evidence of negative impacts (e.g., increases in alcohol or tobacco consumption). The people who receive the one-off transfer, commonly make investments that have a really high return on investment, such as a tin roof which doesn’t need to be rebuilt every year – unlike the thatched roofs they replace. Read the full report by GiveWell.

Schistosomiasis Control Initiative

The Schistosomiasis Control Initiative (SCI) works with governments in sub-Saharan Africa to create and scale up mass deworming programs that rid people of intestinal worms. SCI distributes praziquantel (a drug used to treat schistosomiasis) and albendazole (for hookworm, roundworm, lymphatic filariasis etc) to at risk populations at a cost of $0.68 per person treated.

Data collected by SCI itself as well as two Cochrane Reviews show that this cures most infections, but the evidence is less strong on how this improves people’s lives (trials disagree on the magnitude of the effects). Despite this, some of the effects that have been found for this $0.68/person program look pretty encouraging:

  • 25% increase in school attendance and an improvement on tests of literacy and numeracy
  • 25% increase in income
  • 2.4g/L increase in haemoglobin (Hb) (which carries oxygen – low Hb levels are called anaemia and cause fatigue, weakness, breathlessness)

GiveWell’s full write up

Meta-charitiesGiving What We Can moves between $6 and $60 for each dollar donated to it.

Imagine you’ve made a pie, which you’re about to hand out to your friends. You allocate it according to what will create the most happiness of the group, but because of the pie-to-group ratio, everyone only gets a tiny sliver. What can you do to remedy this? If you get John to make a pie too, you’ve doubled your pie contribution! Finding people to work on some of society’s most pressing issues can be just as useful as John baking another pie.

80 000 Hours, an Oxford careers advisory service, estimated that they coach enough people report a significant career change towards more impactful careers for an average of $7000, including all costs to running an organisation.

GiveWell and Giving What We Can (‘GWWC’) research ways we can give more effectively, and the latter estimates that for every dollar donated to them, between $6 and $60 is donated to effective charities. The Center for Applied Rationality (‘CFAR’) runs workshops teaching participants how to think about problems and how to get things done; so we can get effective things done more efficiently. The Global Priorities Project researches what sort of research needs to be done most urgently.

Non-human animal suffering

A recent consensus among neuroscientists supports the position that many non-human animals have the capacity to suffer. Upon learning this fact, a lot of people become vegetarian or vegan; but what route will lead to the greatest reduction in animal suffering? Becoming a vegetarian limits your meat consumption to near zero, but this doesn’t scale. Online advertising does scale, and appears to influence someone to become a vegetarian for around $12.

Existential risk reduction

So, you know human lives, right? Pretty good. What if they no longer existed? Pretty not good.

Oxford philosopher Nick Bostrom in Superintelligence: Paths, Dangers, Strategies, attempts to instill the value of the lives of future humans. If 10% of the 2*10^20 stars in our Hubble volume were eventually able to be habitable by 1 billion humans, for 1 billion years (one quarter of the time life has been on earth), 10^35 lives are at stake. Representing “all the happiness experienced during one entire such life with a single teardrop of joy, the happiness could fill and refill the Earth’s oceans every second, and keep doing so for a billion billion millennia. It is really important that we make sure these truly are tears of joy.” Imagine we get all this progress on global poverty alleviation and an asteroid wipes all 7 trillion of us. Sup-optimal, really. Even still, the death on this pale blue dot pales in comparison to the loss of life of those 100000000000000000000000000000000000 (10^35) lives. So, the expected value (probability x outcome; a method some people have disagreements with) a reduction in the likelihood of us losing 10^35 future lives by 1 trillionth is still 10^23 lives.

I mentioned an asteroid because it’s pretty easy to intuitively grasp the danger of it. However, Bostrom and most others in the field are mainly concerned about anthropogenic causes of intelligent life extinction. This is mainly because the base rate for most environmental disasters is really low (one in ten to hundreds of thousands) and it gets a lot of research funding already, so is less neglected. In the asteroid example, NASA et al. track 95% of near-earth asteroids. So, if not asteroids, then what?

Some commonly cited examples are nuclear war, synthetic biology and pandemics, geoengineering, artificial intelligence, and [unknown]. Few could have predicted the colossal and diverse effects the internet had, let alone before it was even conceptualised. What development lies around the corner that will shape our world, for good or bad?

Some organisations working on this doing research and policy include:

Further reading

GiveWell on Global Catastrophic Risks | 2
Nick Beckstead’s PhD Thesis – On the Overwhelming Importance of Shaping the Far Future
Max Tegmark on “The Future of Life: a Cosmic Perspective”

Did I miss something? Let me know in the comments or send me a message.