If I think AMF is the best donation opportunity, but donating to The Humane League is better than going vegetarian (because it would be very cheap to “offset” my diet), it shows that donations to AMF are very much better than going vegetarian. This shows going vegetarian makes a small contribution to my potential social impact, so I shouldn’t do it unless it involves negligible sacrifice.
Does it actually show this? I generally hear the argument go something like this:
You can probably convert a lot of vegetarians by donating to The Humane League, which is better than becoming vegetarian yourself. Therefore donating to THL is better than being vegetarian.
Naive estimates say THL does more good than AMF, but AMF has much more robust evidence than THL, so donating to AMF is better.
Therefore donating to AMF is better than being vegetarian.
Parts 1 and 2 use contradictory claims. Part 1 claims that naive expected value dominates, and part 2 claims that robustness of evidence dominates.
Michael, do you have an example? I’ve never seen the union of those 3 in one argument before, although I have seen each of the three claims made by different people.
E.g. it doesn’t describe this post by Jeff Kaufman or this by Greg Lewis. The usual reasons I hear from such people favoring AMF over THL are greater flow-through effects or lower weight on nonhuman animals.
Separately, I hear people, e.g. Tom Ash and Peter Hurford, saying something like #2, but they are themselves vegetarian, and not making arguments for offsetting that I have seen. Indeed, they have challenged it on the basis that the estimates for ACE charities are not robust, which is consistent and contra the argument you described.
I hear people separately make #1 and #2, I can’t recall hearing someone say both #1 and #2 in a single breath. But if you favor AMF over THL because AMF has stronger evidence behind it, that doesn’t preclude going vegetarian. “AMF is better than THL” is not a good argument against being vegetarian, and doesn’t show that vegetarianism is negligible compared to AMF donations, which is the argument Ben was quoting.
So you don’t actually hear people making the argument you mentioned, and the published arguments by Kaufman and Lewis don’t suffer from the inconsistency you mention? Kaufman makes an argument that counting human and cow lives equally, modest AMF donations can be a bigger deal than dairy consumption, while Lewis argues that if one takes ACE estimates seriously, then modest donations to ACE-recommended charities can be a bigger deal than general carnivory.
On the question of donations to AMF vs THL, Kaufman weights AMF over ACE charities because he cares less about nonhuman animals than humans. Some others do so because of flow-through effects. Lewis is vegetarian, but I think mainly donates to poverty and existential risk related things, and I don’t know his precise reasons but they aren’t germane to his essay.
“is the argument Ben was quoting.”
Ben’s description didn’t specify someone thinking AMF was better because they didn’t believe in the robustness of THL ‘animals spared’ estimates. You inserted that, which created the tension in your hypothetical argument. People who favored AMF over THL because of flow-through effects, or because of weighting humans more, wouldn’t have that tension (I would argue the flow-through view would create other tensions, but that’s a different story).
Does it actually show this? I generally hear the argument go something like this:
You can probably convert a lot of vegetarians by donating to The Humane League, which is better than becoming vegetarian yourself. Therefore donating to THL is better than being vegetarian.
Naive estimates say THL does more good than AMF, but AMF has much more robust evidence than THL, so donating to AMF is better.
Therefore donating to AMF is better than being vegetarian.
Parts 1 and 2 use contradictory claims. Part 1 claims that naive expected value dominates, and part 2 claims that robustness of evidence dominates.
Michael, do you have an example? I’ve never seen the union of those 3 in one argument before, although I have seen each of the three claims made by different people.
E.g. it doesn’t describe this post by Jeff Kaufman or this by Greg Lewis. The usual reasons I hear from such people favoring AMF over THL are greater flow-through effects or lower weight on nonhuman animals.
Separately, I hear people, e.g. Tom Ash and Peter Hurford, saying something like #2, but they are themselves vegetarian, and not making arguments for offsetting that I have seen. Indeed, they have challenged it on the basis that the estimates for ACE charities are not robust, which is consistent and contra the argument you described.
You’re correct that Tom and I both assert something along the lines of #2 but have never argued #3.
I hear people separately make #1 and #2, I can’t recall hearing someone say both #1 and #2 in a single breath. But if you favor AMF over THL because AMF has stronger evidence behind it, that doesn’t preclude going vegetarian. “AMF is better than THL” is not a good argument against being vegetarian, and doesn’t show that vegetarianism is negligible compared to AMF donations, which is the argument Ben was quoting.
So you don’t actually hear people making the argument you mentioned, and the published arguments by Kaufman and Lewis don’t suffer from the inconsistency you mention? Kaufman makes an argument that counting human and cow lives equally, modest AMF donations can be a bigger deal than dairy consumption, while Lewis argues that if one takes ACE estimates seriously, then modest donations to ACE-recommended charities can be a bigger deal than general carnivory.
On the question of donations to AMF vs THL, Kaufman weights AMF over ACE charities because he cares less about nonhuman animals than humans. Some others do so because of flow-through effects. Lewis is vegetarian, but I think mainly donates to poverty and existential risk related things, and I don’t know his precise reasons but they aren’t germane to his essay.
“is the argument Ben was quoting.”
Ben’s description didn’t specify someone thinking AMF was better because they didn’t believe in the robustness of THL ‘animals spared’ estimates. You inserted that, which created the tension in your hypothetical argument. People who favored AMF over THL because of flow-through effects, or because of weighting humans more, wouldn’t have that tension (I would argue the flow-through view would create other tensions, but that’s a different story).
I think you’re right actually. A lot of people who prefer AMF to THL are still vegetarian, and that’s totally reasonable and self-consistent.