Where I'm giving and why: Will MacAskill

Summary: I donated this year for two different reasons. I donated to AMF, SCI, Deworm the World, Project Healthy Children, and the Copenhagen Consensus Center because I want to demonstrate my support for extremely cost-effective first-order charities. But I think that the highest-expected-value donation opportunities are to 'meta' organisations. I think that the Centre for Effective Altruism (CEA), Centre for the Study of Existential Risks (CSER), and setting up an Effective Altruism Fund are the three big contenders for best expected bang per buck. Even after trying to correct for obvious personal bias, I think that CEA wins out for my comparatively small donation; if I had info that the relevant position at CSER wouldn't be funded anyway, and if I had more to give (e.g. ~$70k) then I think that CSER would be better.

0. Introduction

I divided my donations into two classes.

First, because my donations partly represent the Centre for Effective Altruism as a whole, I use about half of my donations as a way of demonstrating my support for highly cost-effective organisations that lie outside the core effective altruism community. This year I donated to AMF, SCI, Deworm the World, Project Healthy Children, and the Copenhagen Consensus Center. (Because these have the purpose of showing my support, the usual arguments in favour of giving to just one charity don’t apply). These were all recommended by GWWC except the Copenhagen Consensus Center – I chose to donate to them because I think that promoting global prioritisation is one of the most important activities to be doing, and, though harder to be sure about, their influence is potentially measured in the hundreds of millions or billions of dollars. (My main worry with them is that Lomborg’s (inaccurate) reputation as a climate skeptic might taint the idea of global prioritisation.) I didn’t donate to GiveDirectly because I don’t find GiveWell’s arguments for their ‘upside’ potential to be compelling as a reason to donate on the margin. (Some discussion here, though this post is now a year old).

The second half of my donations I used in whatever way I believe will maximise expected impact, given my particular values and knowledge. I think there are three main contenders.

1. Donating unrestricted to the Centre for Effective Altruism

I think that building the effective altruism movement is in general the highest-value near term activity, for two reasons:

1. Leverage. $1 donated to movement-building generates significantly more than $1’s worth of money and time for the highest-value causes. My view is that this rate of return is far beyond what one could get through financial investment. I’ve written about this before here.
2. Keeping options open. Effective altruists’ donations and use of time are sensitive to the best evidence. I expect our evidence about the best uses of time and money to get much better over the next few years.  This would motivate saving and giving later  (see here for more arguments on either side) – but donating to movement-building is a way of doing that, except with a higher rate of return.

The best precise donation opportunity in this area, in my view, is to CEA, unrestricted such that it can be spent on projects other than Giving What We Can and 80,000 Hours. There are a number of new projects that are plausibly even higher-impact, on the margin, than GWWC/80k, but which don’t have a track record, and so are comparatively difficult to fundraise for. For example, we’ve recently been involved in a number of meetings with people in the UK government, and we wish to develop ideas from CEA and FHI into policy-relevant documents. We are hiring an Oxford academic to help with this (among other things).

CEA unrestricted is where I ultimately chose to give. There is obviously a great risk of bias: it seems to me to be an awful coincidence if I think that the organisation I cofounded is the most cost-effective use of money on the margin. But, even after trying to correct for these biases, CEA remained my top choice. (A partial response to the ‘coincidence’ worry is that a very small number of organisations have ever been started with the aim of maximising the good they do, where the founders’ view of ‘maximising the good’ is the same as or similar to mine (i.e. understood in effective altruist terms).)

2. Donating to the Centre for the Study of Existential Risks

From what I know, this seems like an opportunity for great leverage. Marginal donations will be used to fund program management and grantwriting, which would turn ~$70k into a significant chance of ~$1-$10mn, and launch what I think might become one of the most important research institutions in the world. They have high profile people on the board, Cambridge affiliation, and an already written previous grant proposal that very narrowly missed out on being successful (for ~$10mn). It seems to me that they just need a bit of start-up money in order to hire someone to apply for major grants.

I’ve only been thinking of them as a donation opportunity recently, and would need to know more about CSER’s funding situation before being confident in the above views. For example, if Jaan Tallinn is already paying for a program manager or grant writer, then this leverage opportunity might already have been taken. This expenditure is also pretty lumpy, and I don't expect them to get all their donations from small individual donations, so it seems to me that donating 1/50th of the cost of a program manager isn't as good as 1/50th of the value of a program manager. For those with a larger amount to give, the situation is different.

3. Setting up an Effective Altruism Fund

The idea behind this is that it’s meant to be strictly better than saving and donating later. The idea is that people who are saving in order to donate later tell some central body – the EA fund – how much they’re saving. The EA fund can then advertise the fact that it has some significant amount of money that’s waiting for a high-impact enough opportunity. This could have the following benefits:

1. Incentivises new effective altruism start-ups, which could get seed funding from it (e.g. Effective Fundraising needed just a few thousand dollars – a use of money that I think was clearly worth it).
2. Can be used as loans e.g. for people who want to pursue earning to give but need to retrain.
3. Can be used as ‘emergency reserves’ for several organisations. (Insofar as risks of running out of funding should be only partially correlated between different EA organisations, it makes sense for there to be a shared pool of funding, rather than each organisation having their own reserves.  This means you need a lower total amount of reserves than if every organisation had their own).
4. Can be used as insurance for individuals. Insofar as individuals will be more risk-averse with respect to money when thinking about their own self-interest than when thinking altruistically, we should expect people to be biased in the direction of playing it safe in their career choices. This could be a way of encouraging people to take more risks – by giving them a safety net if things turn out badly.   (Note: though this makes sense in theory, I haven’t seen examples of it in practice, perhaps because there are few paths that result in total failure if they don’t work out. It might apply for someone, for example, thinking about entrepreneurship versus consultancy and worried by the failure rate of new start-ups.)

My view is that this is better than saving and donating later (if you have donations greater than, say, $20,000). But I don’t think it’s as good, right now, as CEA or CSER donations.

GiveWell would have been on the above list of contenders if their funding needs were greater. I think they’re right to want to diversify their funding. But being the very best place to donate is a high bar, and donating to extend GW’s reserves, when they already have GoodVentures as a fall-back, seems unlikely to me to be the best use of funds. Since their first blog post asking for donations, they’ve already raised $250,000; I expect them to soon raise enough to cover next year’s expenses. (Or to figure out a way of being funded by GoodVentures without the problems). As a comparison: whereas GiveWell is looking to raise enough reserves to cover 24 months (in line with their excess assets policy; this includes future pledged donations), because of rapid growth 80,000 Hours and Giving What We Can have recently been operating on only a few months of reserves.