Thanks for the question! I generally believe that it is hard/​impossible to reliably compare CEAs done using different methodologies and approaches. For example, Saulius’ CEA has a different goal than ours and takes into account the overall, average cost-effectiveness of allhistorical work on cage-free campaigns. In contrast, we look at themarginal, future cost-effectiveness of a feed fortification ask. Naturally they will differ a lot. I would expect that marginal cage-free $ would be lower impact than average historical cage-free $.
It’s more informative to compare CEAs within the same methodology, and this intervention is one of the most cost-effective we found. For our future research, we plan to estimate the value of marginal $ spend on an additional cage-free/​broiler campaign to have a better sense of the counterfactuals.
That said, to try to answer your questions, I’ll make a back-of-the-envelope calculation (BOTEC):
If I read correctly Saulius estimated that outside of the US 63-210M hens were affected by cage-free pledges for an average of 15 years of impact with a follow-through rate of 76% for an average = 48M-170M hens helped. He estimated that all global cage-free work costs $36-84M. I didn’t find in his estimates how much of that amount was outside the US. Let’s assume that the cost outside the US was half the total cost, so $18M-42M.
Let’s assume that both the impact and the cost are distributed equally among 7 major orgs. That gives us an effect of 6.8M-24M hens helped for $2.6-6.1M per org. That assumes an impact over 15 years. In our CEA we assume 10 years, so let’s adjust the impact for 10 instead of 15 years – that will be 4.5M-16Mhens per org (6.8M-24M*0.66).
That gives us BOTEC CEA of 1-4.5hens per $, compared to our estimate of approx. 1.8 hens per $. It is generally true that the more extensive the CEA, the lower the estimated cost-effectiveness. So given that BOTEC and our prior research I believe that feed fort alone will be comparable with cage-free, and feed fort + cage-free (something we suggest for India) could have ~twice the impact.
Thanks for sharing and all of your great work! I still find estimating and comparing cost-effectiveness difficult to grasp, but, roughly, how would you say the potential cost-effectiveness of such an intervention (i.e. 12 welfare points per dollar) compare to that of cage-free campaigns (as defined in https://​​forum.effectivealtruism.org/​​posts/​​L5EZjjXKdNgcm253H/​​corporate-campaigns-affect-9-to-120-years-of-chicken-life )? Thanks!
Thanks for the question! I generally believe that it is hard/​impossible to reliably compare CEAs done using different methodologies and approaches. For example, Saulius’ CEA has a different goal than ours and takes into account the overall, average cost-effectiveness of all historical work on cage-free campaigns. In contrast, we look at the marginal, future cost-effectiveness of a feed fortification ask. Naturally they will differ a lot. I would expect that marginal cage-free $ would be lower impact than average historical cage-free $.
It’s more informative to compare CEAs within the same methodology, and this intervention is one of the most cost-effective we found. For our future research, we plan to estimate the value of marginal $ spend on an additional cage-free/​broiler campaign to have a better sense of the counterfactuals.
That said, to try to answer your questions, I’ll make a back-of-the-envelope calculation (BOTEC):
If I read correctly Saulius estimated that outside of the US 63-210M hens were affected by cage-free pledges for an average of 15 years of impact with a follow-through rate of 76% for an average = 48M-170M hens helped. He estimated that all global cage-free work costs $36-84M. I didn’t find in his estimates how much of that amount was outside the US. Let’s assume that the cost outside the US was half the total cost, so $18M-42M.
Let’s assume that both the impact and the cost are distributed equally among 7 major orgs. That gives us an effect of 6.8M-24M hens helped for $2.6-6.1M per org. That assumes an impact over 15 years. In our CEA we assume 10 years, so let’s adjust the impact for 10 instead of 15 years – that will be 4.5M-16Mhens per org (6.8M-24M*0.66).
That gives us BOTEC CEA of 1-4.5hens per $, compared to our estimate of approx. 1.8 hens per $. It is generally true that the more extensive the CEA, the lower the estimated cost-effectiveness. So given that BOTEC and our prior research I believe that feed fort alone will be comparable with cage-free, and feed fort + cage-free (something we suggest for India) could have ~twice the impact.
This is great stuff, Karolina! Thanks for sharing your thoughts on this.