[W]hile the problem of interpersonal comparability of utility is a tricky one, it is not insoluble in principle. It is conceivable that, perhaps several hundred (or a thousand) years from now, neurology may have advanced to the stage where the level of happiness can be accurately correlated to some cerebral reaction that can be measured by a ‘eudaimonometer’. Hence the definition of social welfare [in terms of the sum total of individual happiness] is an objective definition, although the objects are the subjective feelings of individuals.

The practical reality is that existing cost-benefit analyses of animal welfare policies are speciest: they only explicitly consider the benefits and costs of the policy to people.