Yesterday, I published EAs should donate 2% to warm fuzzy causes and 8% to EA causes. It's just a brief argument in favor of the claim I make in the title.

This post is the least popular I've ever written, yet I continue to think it's a great idea. This motivates me to keep writing about and promoting it. I think the bad reception points to a significant discrepancy between my values or worldview and that of the rest of the forum, which seems important to dig into.

Step 0: clarifying the proposal

Not all EAs believe in having something like a "standard" one must achieve to count oneself as an EA. Others think this is pretty important, and that taking the Giving What We Can (GWWC) pledge is a good norm for what that standard ought to be, or is already that de facto standard.

I'm not actually trying to argue this point here, although I do have an opinion. I'm arguing that if we have something like the GWWC pledge as a standard, then it should be the modified 2%-fuzzies/8%-EA-causes pledge (or an individualized variant like 2%-fuzzies/10%-EA-causes). I didn't make this clear in the last post, but for the comments section here, please do not argue over whether or not we ought to have a pledge-like standard at all - it's a good topic for discussion, but not the focus of this post.

Step 1: reading the tea leaves of karma and comments

Based on 13 people's worth of karma and a few comments, I'm not confident on why it attracted so many downvotes.

One comment suggested this is because it seemed to conflate the Giving What We Can pledge with being in EA, or to promote "shoulding:"

I don't think taking the GWWC pledge should be a prerequisite to consider yourself an EA (which, it's not a prerequisite now). If your post had said "GWWC members should..." or "EAs who donate 10% should..." instead of "EAs should..." then I wouldn't have disagreed with the wording.

But another even more highly upvoted respondant disagreed:

I think donating at least 10% of one's income per year should be a norm for any person who identifies as part of the EA community, unless doing so would cause them significant financial hardship.

Another suggested that EA is, on some level, incapable of doing very much thinking about perceptions without undermining our focus on doing good effectively:

Roughly, I think the community isn't able (isn't strong enough?) to both think much about how it's perceived and think well or in-a-high-integrity-manner about how to do good, and I'd favor thinking well and in a high-integrity manner.

They also claim that:

donating for warm fuzzies is generally an ineffective way to gain influence/status.

This comment got the most agreement karma (from 7 people), so I'm guessing this is probably why several disagreers disagreed.

Step 2: arguing against myself

If you believe GiveWell's 2020 estimate that the cost to save a life is $3,000-$5,000, then a harsh (but maybe true) responds to my post is that if 2-3 $80,000/year earners took my advice to donate $1,600 to ineffective charities, then my advice would have annually killed somebody who'd otherwise have lived. If you believe there are bigger returns to investment in other even more effective charities, this bites even harder.

This is true as far as it goes, but it's also true that putting the GWWC pledge at 10% instead of 12-13% has the same effect. If it's OK to put the GWWC pledge recommended amount at 10% instead of 12%, but not to shift to 2%-fuzzies/8%-EA-charities, then an argument based on the individual donor's individaul impact on lives saved or not saved doesn't make sense.

One other way to attack my 2%-fuzzy/8%-EA pledge idea while defending a 10%-EA-pledge is to appeal either to the importance of having a rock-solid, immutable standard, or an idea along the lines that the 10%-EA pledge is almost perfectly calibrated to maximize overall donations. I've never heard the latter idea argued before and don't see any reason to think it is true, so I'm going to set it aside.

Rock-solid, simple standards are potentially nice because there's the potential to build common knowledge and a sort of "brand" around the pledge. 10% may not be perfectly calibrated to the percent, but it's a nice round number that's also substantial enough to make a big difference in the world and represent a very substantial commitment to putting your money where your mouth is. 2%/8% split across two different priorities is a lot more complicated. Maybe that makes it worse as a standard and we really need a very simple, clear, immutable standard?

I have met a couple EAs who seem sort of tortured by this dilemma where they feel like they ought to give hugely more of their money or time and energy to an EA cause, and feel like they've fallen horribly short for failing meet this harsh standard. And so part of the idea of a 10% standard might be to tell them something like this:

Nobody questions your sincere desire to do a lot more for the world, or your ability to do so, and at the same time, nobody else here thinks you're morally obligated to do more than meet this 10% standard. And not to make you feel even worse, but if this movement is causing people to suffer by exacerbating their moral perfectionism, that just seems like a bad think for the cause as a whole. We need people to join EA and be productive and donate to good causes, and if it's full of people suffering from moral perfectionism, that's probably going to interfere with the good we can do as a movement. So try to feel good about meeting but not exceeding that 10% standard, unless you can do so in a way that feels lighthearted.

Usually, if EAs are arguing against a 10%-EA-pledge standard, they're arguing in favor of more donations to EA causes, not less. Will MacAskill said in an interview that he thinks it's morally obligatory for the very wealthy to donate enormous percentages of their income to effective charities. This is the straight-line computation once you start thinking about the idea of earning to give to effective charities; the 10% standard we is layered on top of that, and usually defended as a way to keep the movement human-compatible.

So it seems like if you think the 10%-EA standard is fine but the 2%-fuzzies/8%-EA standard should be downvoted to obscurity, then it might be because you think it's important we not accept further concessions of effective giving in exchange for human compatibility or whatever reputation-bolstering the change would achieve for the movement. Maybe you think we really ought to be giving more than 10%, and you're accepting a compromise at 10%, but you don't want to go lower?

Let's also explore a criticism made in a comment on my previous post:

Roughly, I think the community isn't able (isn't strong enough?) to both think much about how it's perceived and think well or in-a-high-integrity-manner about how to do good, and I'd favor thinking well and in a high-integrity manner.

I'd guess donating for warm fuzzies is generally an ineffective way to gain influence/status. 

Here, it's not the 2%/8% rule itself that is being criticized, but the motivation for the rule, which is to improve the (mainly external-facing) reputation of the EA movement. "Make whatever pledge you want, but don't put a lot of your mental energy into thinking of ways to bolster EA's reputation. Insofar as you want to do good, your reputation-bolstering thoughts will detract from that."

Let's give this due consideration.

I think it's pretty clear that in a year of EA scandal, the worst was the FTX crisis. And before Sam Bankman Fried's downfall, he had already made EA look bad by generating news stories along the lines of "An EA crypto billionaire is trying to buy a congressional election." Here, the problem was reputation-damaging donations (and, to be clear, so wasteful in retrospect as to cancel out many entire lifetimes of donations by average people, to the tune of about 2,000 not-saved lives by GiveWell's estimates).

It's clear that a bad donation like this one can damage reputation. But the above comment claims that making for-the-fuzzies donations (i.e. to a conventional, not-EA charity) does not have an effect in the opposite direction.

Does that mean you can only lose or gain status and reputation via making EA-type donations, while reputation isn't affected by for-the-fuzzies donations? I don't think this is true - plenty of billionaires get lots of positive press for giving away huge amounts of money to causes most EAs would probably consider not to be cost-effective.

Does that mean your reputation can be only damaged or unaffected by a donation, but not improved? We have evidence against this - the reputation of formerly widely reviled billionaires, like Bill Gates, was dramatically improved by their philanthropic efforts. We also know that certain political figures, like Bernie Sanders, gain a sense of authenticity when they're largely or entirely funded by masses of small donations.

I'm left to think that the argument must be that shifting to a 2%-fuzzies donation won't fundamentally change the way EA is perceived, but will  result in the loss of real counterfactual good through the lost charitable work that the extra money could have accomplished. And that there's really no margin on which we'd be able to build EA's reputation by donating to non-EA causes in a way we'd be comfortable with.

This in turn might be because, although making large donations has improved the reputation of quite a few billionaires, it's not the donation amount or proportion that really matters - it's the way the billionaire is able to generate positive press coverage via philanthropy. From this point of view, merely slightly adjusting the proportion we donate to EA causes won't generate the kind of positive press that would lead to a bolstered reputation.

I think these are three good hypotheses for why people might disagree with me:

• The simplicity and constancy of the current 10%-to-effective-charities Giving What We Can pledge makes it the best standard for us to have.
• We ideally would donate more than 10%, and all our donations ought to be to EA causes, but we must accept a grudging 10% compromise for the sake of human compatibility. We don't want to encourage going any lower.
• Shifting to a 2%/8% fuzzies/EA pledge standard will represent real loss of effective charitable resources, but will fail to improve EA's reputation. Maybe there's a way to bolster EA's reputation, but this isn't it.

Addressing these potential criticisms

0%/10% optimizes for simplicity and constancy, 2%/8% optimizes for playing nice with others; I think EA's reputation needs the latter more than the former. 

N of 1, but I am personally more comfortable with a 2%/8% standard than a 0%/10% standard. My family does talk about which organizations we donate to. I expect that being able to say that EA encourages me  to donate 2% to local causes and 8% to things like AI risk, pandemic prevention, malaria and so forth would genuinely smooth out those conversations and lead to greater appreciation of the movement. And I do suspect that a movement that encourages 2% donations to any charity and 8% to EA causes would be perceived as "playing nicer" by non-EAs. It's a strange psychology but I think a certain zero-sum mentality around charity is real, as evidenced by a lot of the hit pieces written about EA by organizations like Charity Navigator.

A lot of EA's reputational damage comes from the many ways we diverge from the norm culturally (polyamory), intellectually (focus on cost/benefit analysis), socially (Silicon Valley links) and charitably (prioritizing global and X-risk-related causes) seem to threaten bedrock institutions of society (with things like election-buying) and values (like an appreciation of "heart" in the charitable space). Making concessions to and participating in normalcy with a 2%/8% standard seems like a good way to start addressing some of these concerns, though certainly not all of them.

The 10%-is-a-grudging-compromise issue is logically consistent, but on its own it fails to value the reputational benefit I perceive in the 2%/8% standard.

We have a standard that encourages EAs to live well instead of donating until they themselves are in poverty. The reason we have that is partly out of a sort of moral parliament concern for the idea that we ought not to extract all resources from EA, even if a pure utilitarian calculation would say we should. But a bigger reason is that we think it's impractical - who would want to participate in such a demanding movement? The EA charity world gets a smaller slice of a much bigger pie when we limit people's commitment in this way. 

Shifting to a 2%/8% standard is just following the same sort of logic. It makes it more possible for people with at least partial commitments to non-EA causes and values to participate or at least to appreciate EA. I think this enhances the "smaller slice of a bigger pie" effect even further.

Would a 2%/8% standard be ineffective in bolstering EA's reputation and represent a simple waste of resources?

As I've argued above, I just disagree that it would be intrinsically ineffective. It could be ineffective if executed in the wrong way. An anonymous donor doesn't gain reputation points; a known donor often does as long as they can frame the donation in a way that others view in a positive light. So a simple shift to a 2%/8% donation isn't the end of the strategy - we'd also have to figure out how to frame this new direction in a way that others approved of. Certainly not everybody would care, and some people would frown on it, but I think it's easier to get people to appreciate this sort of balanced approach, and too easy to frame a 0%/10% standard as "extreme."

So far, I stick with my advocacy of a 2%-to-warm-fuzzy-charities, 8%-to-EA-causes Giving What We Can-style pledge as superior for the health of the EA movement to the current conventional 0%/10% pledge. I am happy to debate this in the comments, and plan to publish more on this in the near term future.