[JamesBarney]

The whole point of EA is to move toward a system where charities are judged by the good they do and not their marketing campaign.

AMF and other top givewell charities seem far more focused on the good they can do than more classic charities like WWF or Plan Parenthood.

[emodendroket]

I am rather skeptical of the notion that this can be meaningfully quantified.

[JamesBarney]

Surely you think between two charities, one that takes a $1,000 and buys caviar for starving children and one that feeds 1000x as many bread the second one does more good per $ than the first.

[emodendroket]

Sure, it is easy to invent fantastically silly uses of money. Is that what is at issue?

[ben_w]

Kinda, yes.

It's certainly the scale of impact difference between some of the charities:

Charities can be useless or actively harmful: https://www.givewell.org/international/technical/criteria/im...

vs.

$4500/life saved: https://www.givewell.org/our-giving-funds#Top_Charities_Fund

They also say why it's that high: https://www.givewell.org/impact-estimates

Another example I've heard is hospitals asking people to stop skydiving to raise money for them, as the mean cost of medial care the average skydiver needs due to skydiving injuries exceeds the average money raised by skydiving.

[woodruffw]

And what theory of Good determines how we do that judging? The different "schools" of EA seem to have radically different approaches ("longtermism" versus maximizing QALYs for living beings).

And even this punts on perceptive utility, the kind that Nozick warns us about[1]: it's entirely possible that Johnny in Country X gets more utility out of $100 in charitable giving than 50 others in Country Y. But this seems like a really bad logical consequence, the kind that the Internet Rationalist EA community uses to justify spending money on AGI research instead of donating to AMF.

Overall, it's much easier to live in a world where we can determine that some set of charities are worth giving to, and not try too hard to order between them.

[1]: https://en.wikipedia.org/wiki/Utility_monster

[JamesBarney]

Do you think it's good to donate to charities?

Do you think that some charities are more effective than other ones?

Do you think we should try to measure that to determine which ones are more effective?

[emodendroket]

When one considers that there are numerous charities with exact opposite goals I’m not sure how you can consider your three questions to have self-evident answers.

[JamesBarney]

Most charities don't have a charity with an exact opposite goal.

Most charitable donations are net good. Sure there exist a small subset of charities that aren't probably doing harm but that's the exception.

Would you argue that charitable giving is neutral or bad?

[emodendroket]

In some light it could be seen as "bad," at least as currently structured in this country: we basically allow people to withhold tax money and instead direct the money as they see fit, which has an anti-democratic implications -- in effect it as though we went to the richest Americans and said "why don't you decide how a substantial chunk of tax dollars are spent rather than Congress?" Beyond the more philosophical aspects, having been in the sausage factory for a bit, and having read about others, I think it is very easy for charities to just happen to serve their mission in a way that just happens to be more beneficial for their patrons (even if they could put superficially convincing numbers on it to prove they're highly "effective").

But I suppose I would say it's neutral; it could be good or bad depending on various characteristics of the charity itself.

[JamesBarney]

I think on avg. charities aren't so bad that the world is better off if you buy a yacht and give 20% to the government than to donate it to the average charity.

But do you at least agree the government does less good with 20 cents than a givewell charity like the AMF does with a dollar?

[woodruffw]

I think giving to charities is supererogatory, at best. At worst, it’s a reflection of a heteronomous will.