New members thinking that their pledge would be fulfilled by donating to a charity in their local area, rather than considering charities on a global scale.
Unfortunately, it seems that this would satisfy the pledge, assuming the donor genuinely thought those charities were the most effective. And there are plausible epistemic states that might lead to such a view—for example, skepticism about the reliability of reports about the impact of distant actions.
This would not be a concern if GWWC was prepared to say “and your donations have to be to charities we approve of.” Unfortunately, because the pledge also says
most effectively
this would require GWWC to judge that these charities were all equally effective. So, for example, FHI and SCI would have to be equally effective. Which does not sound very plausible—while I can understand thinking that FHI is much better than SCI, or SCI much better than FHI, thinking the two are equal in effectiveness seems hugely implausible. I don’t know of anyone who actually holds that view.
So as far as I can see, either
Donating locally could satisfy the new pledge
GWWC should drop the word ‘most’
All the charities GWWC endorses as pledge-compliant must be exactly equally effective
or GWWC can only recommend one charity (which is actually a sub-case of the previous option).
there are plausible epistemic states that might lead to such a view—for example, skepticism about the reliability of reports about the impact of distant actions. This would not be a concern if GWWC was prepared to say “and your donations have to be to charities we approve of.”
‘Plausible epistemic states’ in the sense of ‘epistemic states someone could plausibly have’ shouldn’t be conflated with ‘plausible epistemic states’ in the sense of ‘well-evidenced epistemic states’. The latter is what matters. If the evidence all suggests that GiveDirectly’s reports are accurate, then it is the responsibility of pledge-takers to take that information into account in selecting their target charity.
The pledge is normative about epistemic methodology; this is the third alternative we miss when we only consider the options ‘anything goes’ and ‘the right charity is whatever GWWC says it is’. A certain minimum threshold of reasonableness is required to follow the GWWC pledge; otherwise absolutely any cause can fit the criterion. But that threshold isn’t so high that everyone who follows it is sure to get the exact optimal answer; and it isn’t fitted such that GWWC is defined as always being in the right.
All the charities GWWC endorses as pledge-compliant must be exactly equally effective
An option you don’t consider is that “I shall give at least ten percent of what I earn to whichever organisations can most effectively use it to help people [in developing countries]” is consistent with ‘We’re too uncertain about which organization is most effective to give solid concise instructions on that front’. Taking the pledge means doing your best (or putting in a nontrivial effort) to figure out the best cause, and then giving to that cause; but as long as you end up with a non-ridiculous candidate, you’ve probably met that standard.
If the evidence all suggests that GiveDirectly’s reports are accurate, then it is the responsibility of pledge-takers to take that information into account in selecting their target charity.
Right. Given the strong mean reversion shown in intervention effectiveness I think it is not totally unreasonable to doubt that evidence. Personally I think GiveDirectly is credible but I don’t think this is the only epistemically justifiable position to take.
Taking the pledge means doing your best (or putting in a nontrivial effort) to figure out the best cause, and then giving to that cause; but as long as you end up with a non-ridiculous candidate, you’ve probably met that standard.
But the pledge isn’t merely to attempt to do so—it is to actually do it!
Unfortunately, it seems that this would satisfy the pledge, assuming the donor genuinely thought those charities were the most effective. And there are plausible epistemic states that might lead to such a view—for example, skepticism about the reliability of reports about the impact of distant actions.
This would not be a concern if GWWC was prepared to say “and your donations have to be to charities we approve of.” Unfortunately, because the pledge also says
this would require GWWC to judge that these charities were all equally effective. So, for example, FHI and SCI would have to be equally effective. Which does not sound very plausible—while I can understand thinking that FHI is much better than SCI, or SCI much better than FHI, thinking the two are equal in effectiveness seems hugely implausible. I don’t know of anyone who actually holds that view.
So as far as I can see, either
Donating locally could satisfy the new pledge
GWWC should drop the word ‘most’
All the charities GWWC endorses as pledge-compliant must be exactly equally effective
or GWWC can only recommend one charity (which is actually a sub-case of the previous option).
‘Plausible epistemic states’ in the sense of ‘epistemic states someone could plausibly have’ shouldn’t be conflated with ‘plausible epistemic states’ in the sense of ‘well-evidenced epistemic states’. The latter is what matters. If the evidence all suggests that GiveDirectly’s reports are accurate, then it is the responsibility of pledge-takers to take that information into account in selecting their target charity.
The pledge is normative about epistemic methodology; this is the third alternative we miss when we only consider the options ‘anything goes’ and ‘the right charity is whatever GWWC says it is’. A certain minimum threshold of reasonableness is required to follow the GWWC pledge; otherwise absolutely any cause can fit the criterion. But that threshold isn’t so high that everyone who follows it is sure to get the exact optimal answer; and it isn’t fitted such that GWWC is defined as always being in the right.
An option you don’t consider is that “I shall give at least ten percent of what I earn to whichever organisations can most effectively use it to help people [in developing countries]” is consistent with ‘We’re too uncertain about which organization is most effective to give solid concise instructions on that front’. Taking the pledge means doing your best (or putting in a nontrivial effort) to figure out the best cause, and then giving to that cause; but as long as you end up with a non-ridiculous candidate, you’ve probably met that standard.
Right. Given the strong mean reversion shown in intervention effectiveness I think it is not totally unreasonable to doubt that evidence. Personally I think GiveDirectly is credible but I don’t think this is the only epistemically justifiable position to take.
But the pledge isn’t merely to attempt to do so—it is to actually do it!