This is more of a side remark, not a direct reply to your comment. I'm a utilitarian for general policy making, especially when statistical data is at play, bu

This is more of a side remark, not a direct reply to your comment. I'm a utilitarian for general policy making, especially when statistical data is at play, but I cannot understand utilitarians who believe that all of morality should be utilitarian. Even elaborate versions are flawed, even e.g. Harsanyi's utilitarianism where you are supposed not to know who you are in society (similar to Rawl's "veil of ignorance" but formally worked out in his framework). It's way too easy to find counter-examples that do not match our moral preconception and it takes a huge leap of faith to just ignore these. For example, it is easy to find real-world and fictitious scenarios in which utilitarianism would tell us to torture or murder someone, yet existing laws and traded morality strongly go against these recommendations. Another huge problem is that utilitarians rarely take into account human agency, but we do that all the time in our lives. There is a gigantic difference between risking your own life and telling someone else to risk his or her life. Most utilitarians don't even try to model these differences, and I personally suspect that there could be an impossibility theorem hiding in versions of utilitarianism that attempt to take agency formally into account. (Not that I have investigated this in detail yet, it's just a hunch.) The biggest flaw of utilitarianism is my view that it oversimplifies values. It is often based on the assumption of value monism but in reality we base our judgments on multiple value categories that are often in conflict with each other. Related to that, all utilitarians I know of suggest compensatory decision methods but there are many good reasons why we should sometimes use noncompensatory outranking methods. The idea is crazy that you can murder someone as long as that allows you to give n people $1 each, yet it is implied by any compensatory value aggregation method.