Economists assume a utility curve in the same way that physicists assume spherical cows. They are a useful abstraction which help explain many things, but don't reflect reality completely.

https://www.econlib.org/library/Enc/BehavioralEconomics.html