Without information how exactly the money will be use, one can consistently argue only for 0% or 100%, because those extremes do not depend on details.
Even if I’m uncertain about how the money would be used, my expected utility as a function of redistribution could still peak somewhere between 0% and 100%.
Assuming that you are super rational and you know the prior probabilities of the possible uses of money, then yes, you could calculate that e.g. a 42% taxation brings the highest expected utility.
I just don’t expect a human to do this correctly, or even approximately correctly. So even if a group of humans will use the same reasoning, they will get widely different results (because their prior probability distributions will be different, their knowledge imperfect, and their calculations incorrect).
What should I have said instead is that only those kinds of reasoning which always give you result 0% or 100% (such as saying that any tax is unacceptably evil theft; or that humans are unable to live without a society, so they owe everything to society) will cause a group of humans that use the same reasoning to consistently produce the same numbers. Also, that kind of reasoning is more simple, and more attractive, so more people will do it. Though, there are also other attractive kinds of reasoning, such as: the same as yesterday; a little more than yesterday; a little less than yesterday.
Even if I’m uncertain about how the money would be used, my expected utility as a function of redistribution could still peak somewhere between 0% and 100%.
Assuming that you are super rational and you know the prior probabilities of the possible uses of money, then yes, you could calculate that e.g. a 42% taxation brings the highest expected utility.
I just don’t expect a human to do this correctly, or even approximately correctly. So even if a group of humans will use the same reasoning, they will get widely different results (because their prior probability distributions will be different, their knowledge imperfect, and their calculations incorrect).
What should I have said instead is that only those kinds of reasoning which always give you result 0% or 100% (such as saying that any tax is unacceptably evil theft; or that humans are unable to live without a society, so they owe everything to society) will cause a group of humans that use the same reasoning to consistently produce the same numbers. Also, that kind of reasoning is more simple, and more attractive, so more people will do it. Though, there are also other attractive kinds of reasoning, such as: the same as yesterday; a little more than yesterday; a little less than yesterday.