The graph of monetary utility may look like a logarithmic graph at first glance, but that's just because it's more like a C1 + (x-C2)^3 graph where you haven't followed x far enough to the right.
Remember when Michael Bloomberg spent 500 million dollars to be on the Democratic debate, and Elizabeth Warren burned that money down with a single zinger?
> We have thus confirmed the (cardinalist) assumption of nineteenth century economists that marginal utility of income declines with income.
https://cep.lse.ac.uk/pubs/download/dp0784.pdf
People say stuff that needs evidence then say "according to economic theory" like its a source.
(fwiw I agree that this isn't a good argument for a log scale)
