How to choose a charity to donate to

So you're going to donate to a charity. When selecting which one to donate to, are you going to simply donate to whichever one happens to ask you first? (Hint: this is probably not going to lead to you finding the best answer!)

A very reasonable solution here would be to outsource that thinking to someone else. You can find some recommendations on the SoGive website.

If you choose to do your own research, here are some factors to take into account:

Factors to consider: Mega-summarised list
---------------------------------------------------------------------------------------

• Cost-effectiveness
• Evidence
• Neglectedness
• Management

Factors to consider: Summarised list

---------------------------------------------------------------------------------------

• ·        Cost-effectiveness, which SoGive's data can help with. This can be broken down into:
• o   How much does it cost for the charity to achieve something
• o   And what does the beneficiary get for your money
• ·        Strength of the evidence: does it actually work? This question is more important than it looks because of funder-beneficiary separation and the optimiser's curse
• ·        Neglectedness: you want to try not to fund something that lots of others are funding anyway. Assessed by considering the amount spent on the cause area and a measure of the prevalence of the issue. Not the most important consideration
• ·        Strength of the management team: you can assess this by meeting the management or finding out from someone else who has. Unless your potential donation is large (e.g. >10% of the charity's annual budget) you will not be likely to get the chance to assess this yourself.

Factors to consider: More details
---------------------------------------------------------------------------------------

Cost-effectiveness:
The logic here is familiar to anyone who has bought something for themselves -- you normally ask how much does the thing cost, and what do you get for your money. To find the equivalent information for charities, you can try looking for it on SoGive's database by searching for the charity on app.sogive.org. If the charity isn't there, then you can request it (e.g. by contacting me directly)

To help you get a feel for whether something constitutes good value for money or not, here's a few examples:

Relatively inexpensive, and therefore probably good value for money:
Solar Aid: <£10 for one person to get access to a solar lamp
Project Healthy Children: <10p for one person to get better nutrition
Against Malaria Foundation: <£10 for one malaria net

Relatively expensive, therefore maybe not good value for money:
Guide Dogs for the Blind: ~£60,000 for one guide dog
Make a Wish: >£5,000 for one child to have a special day
Grief Encounter: >£1,000 for a bereaved family to be supported
Fistula Foundation: ~£1,000 for one fistula repaired (in this you could argue that the probability of having a huge impact on the beneficiary is high enough that this does constitute good value for money)



Strength of the evidence:

On one level, this is obvious. If you're more confident about the evidence, this is better.

However this masks a couple of surprisingly important subtleties:
- a disturbingly large number of charities may be achieving nothing (because of funder-beneficiary separation)
- the very act of choosing the best one may force the importance of evidence (optimiser's curse)

Many charities may be achieving nothing (because of funder-beneficiary separation)

It looks like 75% of social programmes and services achieve small to no effect (or even negative impact) (source: David Anderson, Coalition for Evidence-Based policy)

Why is this?

I believe that the main reason for this is a phenomenon called funder-beneficiary separation. This is a label for the fact that in the world of donor-funded organisations, the person who paid for a thing and the person who benefits from a thing are different people. It means that charities can continue to have very little (or even no or negative) impact as long as they can convince donors to continue funding them.

Do note that the 75% relates to social interventions working directly with people -- something like research to tackle cancer or heart disease isn't part of this. There's also a number of subtleties that I've glossed over, to look into this in more detail, look at this link.

Optimiser's curse

Imagine you're choosing between lots of different things (projects your company could invest in, or -- more germanely -- charitable projects you could fund). You decide to pick the one which is, on average, the best (or has the best "expected" outcome, if you prefer that jargon) Now here comes the counter-intuitive bit...

On average, you should expect the best one to be worse than you expect it to be.

Wait, what?! How can that be? It's because by choosing the one that looks best, you increased the probability that you were picking a project whose benefits were overstated.

A more lengthy and rigorous treatment of this idea can be found in the 2006 paper on the topic.

The way to solve this (also set out in the same paper) is to give more emphasis to those which have better quality evidence backing them.

Neglectedness

It is intuitive that your donation is less likely to be valuable if you're funding something that everyone else wants to fund anyway.

Some things that you can find out to assess this:
- is the charity too rich? (if the charity has mountains of cash already, you probably don't want to fund it; conversely most charities aren't obviously really rich, in which case asking this question won't have told you much) More on this here.
- you could look at how much is spent in total on the cause area and compare that to the prevalence of the problem. If the cause area already has lots of money, the charities within it might be scratching around for good opportunities for what to do (and possibly hitting diminishing marginal returns). In order for this to be useful, you would want to be able to compare this with other cause areas; as far as I'm aware, nobody else has done this sort of data-gathering exercise. If there is enough appetite from our users for this data to be gathered, then SoGive may consider doing this.

Note that even if an area is popular and has lots of people funding it, the specific thing you are funding might be attacking a neglected niche, or it might be an area in which diminishing marginal returns don't apply for some reason, so for this reason I'd suggest we should take it into account as a factor, but not treat it as the most important.

Management

How well a charity executes its mission is important. The best way to assess this is to meet the senior team and get a feel for their competence that way. If your donation is not large (say, <10% of the charity's annual spend) you probably won't get their time (and it's probably not fair to ask). Sometimes there are other ways to find out though:

-- sometimes a charity's leaders have a decent social media profile, and seeing them talk on youtube videos might give you an idea of how they think (this is typically not enough to get a decent picture though)
-- sometimes other charity assessors may have done this assessment -- for example GiveWell includes an assessment of a charity's track record as part of their analysis, however they only assess a small number of charities in depth