Give Collectively

by harsimony3 min read18th Sep 20215 comments


Effective giving

(Cross-posted from my blog)

Charitable giving is one of many great ways to make the world a better place. I have spent some time talking about how we might do it better, but I have largely ignored one major problem with giving. Namely, it is hard for donors to coordinate giving and to take advantage of their collective information.

Consider the individual donor with a small amount to give. Though spreading donations out is one method to ensure that giving is sustainable, it does lower your total charitable impact. In fact, the “optimal” individual donor gives to a single charity! That is, an altruist who wants to do the most good should only give to the charity that they consider to have the highest marginal impact [1].

Contrast this with mega-donors. Since large donors can see diminishing returns when giving to only one charity, it makes more sense to spread donations out. Additionally, since their decisions are more consequential, it makes sense for them to gather more information. On top if this, large donors can also better coordinate with each other, ensuring that they don’t overspend on the same charity. Even better, mega-donors have enough money to start entirely new charitable organizations and get new projects off the ground.

These factors combine to make mega-donation significantly more impactful per dollar than what small donors can achieve.

This is a pretty big limitation on the overall impact of charitable spending, since the majority of charitable spending comes from small donors [2] and the vast majority of potential future donations will come from small donors [3].

We need to find a way to get many small donors to act like a single, large donor, incorporating their combined information, coordinating their giving, and finding new charitable opportunities. In other words, in order to maximize our impact, we need to find a way to give collectively.

Public debate, meta-charities like Givewell, and public charity finances already helps to coordinate giving somewhat, but we can do much better. There are already several promising directions. Donor lotteries take the simple approach of turning many small donors into one large donor by giving all the money to one person. Donation dominant assurance contracts (which I discussed briefly here) are another approach, but while it coordinates giving, it does little to utilize collective information in order to make better decisions. Charity prediction markets make better use of collective information but cannot themselves direct donations.

I want to put forward one new idea to encourage creative thinking in this space. I call it “iterative public donation”. It starts with a group of donors signing a donation-DAC and privately determining how they are dividing their personal budget amongst a list of charities. These budgets are added up and the total donations going to each charity are listed publicly. Next, donors adjust the amount they are giving to each charity based on the current total. This process repeats until the amount donated to each charity equilibrates. This simple mechanism allows donors to respond to how much others are spending and learn others opinions on charity, leading to better coordinated giving overall. This process could be augmented with some form of public discussion and data gathering to improve the knowledge of donors before they participate.

There is plenty of room for new ideas and experimentation here. Finding a way to give collectively would significantly increase the overall effectiveness of donations and permanently change the way we give.


  1. The reason for this is that, for a small donor, their donations have roughly linear impact in the size of donation size. In this case, the impact per dollar is roughly the marginal impact per dollar at the charities current funding level. The situation is different for very large donors, where donations might exhibit diminishing marginal returns, which encourages spreading the donations out.

  2. See the section titled “The 1 percent” at the end of this article.

  3. Personal income in the United States dwarfs total charitable spending, and the combined wealth of the worlds billionaires (that’s $20 trillion per year compared to a one-time $5 trillion, so the net present value is much higher). This means that small donors are essential to increasing total charitable giving in the future.


5 comments, sorted by Highlighting new comments since Today at 4:42 PM
New Comment

I like the way you think! Would you like to be my FB friend? I mostly use FB for discussing ideas.

That said, I didn't find this post so valuable? Like, what's the advantage of doing Iterative Public Donation compared to, say, just donating to any of the EA Funds? Or even a donation lottery?

I agree that the EA funds (and meta-charities like Givewell), are great opportunities to give and can help balance the flow of donations going to different charities. But I don't think that these funds have entirely solved the collective action problem in charitable giving. Rather, they aggregate money from many donors and turn over funding decisions to a handful of experts. These experts are doing great work, and I really respect them, but it doesn't hurt to consider how we might do things even better!

If we really did have a system for small donors to coordinate their giving like large donors, things would look quite different:

  • Collections of small donors would be able to fund specific research projects, found new charitable organizations, and exert significant control over the day-to-day activities of these organizations.

  • Collections of donors would be able to work with mega-donors, governments, and charitable organizations to pursue much larger projects.

  • Collections of small donors would be able to deliberate amongst themselves and make funding decisions based on their combined knowledge.

Charities and EA funds do this in a roundabout way by acting as representatives for many small donors, but this isn't the only way to organize giving. What about a kickstarter for EA research projects? Or a charitable fund where managers are elected by donors? Or a prediction market on how impactful different interventions are? I'm not claiming that these ideas are going to be better than the current instantiation of EA funds, but I want to encourage exploration and experimentation before we settle on these as the only solution to collective donation.

I'm a bit confused about your view here. Why can't EA Funds, with enough money, fund specific research projects and new charitable organizations? Why can't they "work with mega-donors (...) to pursue much larger projects"?

Also, it seems like you have more faith than me in the collective wisdom of many non-experts, compared to a team of experts whose job is to work on the questions full-time.  Do you think a donation would be better allocated by votes from 1000 average EAs who each spend 2 hours of research each, or by a team of 10 highly experienced EAs who each spend 200 hours of research each?

Of course, EA funds can do all of these things, and I appreciate the work they are doing.

I think it is important to be explicit about the structure of EA funds, meta-charities, and charitable foundations: they typically involve pooling money from many donors and putting funding decisions in the hands of a few people. This is not a criticism! It makes a lot of sense to turn these decisions over to knowledgeable, committed specialists in the EA community. This approach likely improves the impact of peoples donations over the counterfactual where people give directly to charities without considering how other are donating.

While I appreciate this system, I don't see why we shouldn't at least consider other systems of collective donation. It seems worthwhile to explore other approaches before settling on one specific model of collective giving.

Also, it seems like you have more faith than me in the collective wisdom of many non-experts, compared to a team of experts whose job is to work on the questions full-time.

Under the right circumstances, many non-experts can and do outperform experts. Tetlock's Superforcasting and prediction markets are good examples of this. That being said, I am highly uncertain as to whether these conditions hold for charitable donation, so experimentation with different funding models seems valuable.

Strongly agree with trying to think about creative ways of effectively pooling wisdom together to make better donations collectively. Prediction markets, as you point out, is an example of a really high-impact related idea, so there might be more in the vicinity. I should've made that clear from the start. I just don't think this particular suggestion (Iterative Public Donation) beats what we've currently got. But, as I've said, I like the way you think. :)