I would agree with you that I am dealing with several highly subjective values. Like I said before, this is just a start. Since I made that video, I have been looking at a possible range of values for valuing life. This of course would start with the use of existing VSL for humans. I will also need to investigate other ways of determining intelligence of animals. It would be good if I could establish a degree of sensitivity to pain for each species. I think investigating a relationship between intelligence and, physical and mental suffering might be a better path to take; this would still be tricky as there are many other factors that are likely to effect pain tolerance which are unique to each species. So to cut a long story short, I would expect a range of values for the value of an animal's life. I might even be able to come up with a distribution of results, wouldn't that be cool.

The original study looked at just the costs of improving conditions on farms to reduce animal suffering and discomfort from existing practices. This approach only captures a very small amount of the costs of cruelty.

What you propose regarding people estimating how they value a human life in relation to a cow's life could produce absolutely any value based on opinion. I think this is very dangerous. What if, like you said, people valued a cow's life at zero. It is better to keep human opinion out of this. Even willingness-to-pay data is based on people's response to various questions and scenarios rather than a direct question on how they value another person's life or their own life for that matter. Brain weight and body weight is not influenced by human opinion and neither is life expectancy. This way we can mostly (there is still my own bias in regards to the selection of criteria to determine the adjusted value of an animal's life, providing a distribution of results should greatly reduce this) remove human bias.

Thank you for the interesting discussion.