Vasco’s analysis implies a much wider welfare range than the one we use.
We’re not confident in our current assumptions, but this is a complicated question, and there is more work we need to do ourselves to get to an answer we believe in enough to act on. We also need to think through consistency with our human-focused interventions.
We don’t use Rethink’s moral weights.
Our current moral weights, based in part on Luke Muehlhauser’s past work, are lower. We may update them in the future; if we do, we’ll consider work from many sources, including the arguments made in this post.
I am a little confused by the above. You say my analysis implies a much wider welfare range than the one you use, but in my analysis I just used point estimates. I relied on Rethink Priorities’ median welfare range for chickens of 0.332, although Rethink’s 5th and 95th percentile are 0.002 and 0.869 (i.e. the 95th percentile is 434 times the 5th percentile).
Are you saying Rethink’s interval for the welfare range of chickens is much wider than Open Phil’s? I think that would imply some disagreement with Luke’s guess. Following his 2017 report on consciousness and moral patienthood, Luke guessed in 2018 a chicken life-year to be worth 0.00005 to 10 human life-years (“80% prediction interval”; upper bound 200 k times the lower bound). This interval is 11.5 (= (10 − 0.00005)/(0.869 − 0.002)) times as wide as Rethink’s on a linear scale, 2.01 (= ln(10/0.00005)/ln(0.869/0.002)) times as wide on a logarithmic scale, and Luke’s interval respecting the 5th and 95th percentile would have been wider. Rethink’s and Luke’s intervals are not directly comparable. Rethink’s refers to the welfare range, whereas Luke’s refers to moral weight. However, I would have guessed Open Phil’s interval for the moral weight to be narrower than Open Phil’s interval for the welfare range, as Open Phil’s funding suggests a comparatively low weight on direct hedonic effects (across animal and human interventions). In any case, I must note Luke was not confident about his guesses, having commented around 1.5 years ago that:
Since this [this] exercise is based on numbers I personally made up, I would like to remind everyone that those numbers are extremely made up and come with many caveats given in the original sources. It would not be that hard to produce numbers more reasonable than mine, at least re: moral weights. (I spent more time on the “probability of consciousness” numbers, though that was years ago and my numbers would probably be different now.)
On the other hand, multiplying Luke’s numbers by the ratio of 10 k between the cost-effectiveness of corporate campaigns and GiveWell’s top charities for a moral weight of 1 shared in the worldview diversification 2016 post[1], one would conclude corporate campaigns for chicken welfare to be 0.5 (= 0.00005*10000) to 100 k (= 10*10000) times as cost-effective as GiveWell’s top charities. In my mind, the prospect of corporate campaigns for chicken welfare being much more cost-effective at increasing welfare than GiveWell’s top charities should have prompted a major investigation of the topic, and more transparent communication of Open Phil’s prioritisation decisions.
“If you value chicken life-years equally to human life-years, this implies that corporate campaigns do about 10,000x as much good per dollar as top charities. If you believe that chickens do not suffer in a morally relevant way, this implies that corporate campaigns do no good”.
Thanks for the feedback, Emily!
I am a little confused by the above. You say my analysis implies a much wider welfare range than the one you use, but in my analysis I just used point estimates. I relied on Rethink Priorities’ median welfare range for chickens of 0.332, although Rethink’s 5th and 95th percentile are 0.002 and 0.869 (i.e. the 95th percentile is 434 times the 5th percentile).
Are you saying Rethink’s interval for the welfare range of chickens is much wider than Open Phil’s? I think that would imply some disagreement with Luke’s guess. Following his 2017 report on consciousness and moral patienthood, Luke guessed in 2018 a chicken life-year to be worth 0.00005 to 10 human life-years (“80% prediction interval”; upper bound 200 k times the lower bound). This interval is 11.5 (= (10 − 0.00005)/(0.869 − 0.002)) times as wide as Rethink’s on a linear scale, 2.01 (= ln(10/0.00005)/ln(0.869/0.002)) times as wide on a logarithmic scale, and Luke’s interval respecting the 5th and 95th percentile would have been wider. Rethink’s and Luke’s intervals are not directly comparable. Rethink’s refers to the welfare range, whereas Luke’s refers to moral weight. However, I would have guessed Open Phil’s interval for the moral weight to be narrower than Open Phil’s interval for the welfare range, as Open Phil’s funding suggests a comparatively low weight on direct hedonic effects (across animal and human interventions). In any case, I must note Luke was not confident about his guesses, having commented around 1.5 years ago that:
On the other hand, multiplying Luke’s numbers by the ratio of 10 k between the cost-effectiveness of corporate campaigns and GiveWell’s top charities for a moral weight of 1 shared in the worldview diversification 2016 post[1], one would conclude corporate campaigns for chicken welfare to be 0.5 (= 0.00005*10000) to 100 k (= 10*10000) times as cost-effective as GiveWell’s top charities. In my mind, the prospect of corporate campaigns for chicken welfare being much more cost-effective at increasing welfare than GiveWell’s top charities should have prompted a major investigation of the topic, and more transparent communication of Open Phil’s prioritisation decisions.
“If you value chicken life-years equally to human life-years, this implies that corporate campaigns do about 10,000x as much good per dollar as top charities. If you believe that chickens do not suffer in a morally relevant way, this implies that corporate campaigns do no good”.