In fact, economists perform a sleight of hand here. While granting that assumptions are not realistic, and are made only ‘as if’, economists next give excuses why predictions cannot be rigorously tested, and then fall back on the theoretical assumptions anyway, no longer only ‘as if’ they apply, but also in deriving policy implications, claiming that markets should be left alone because they are efficient. But that remains to be demonstrated. The snake bites its tail. At first, markets are taken only ‘as if’ they are efficient, this is never convincingly corroborated, and then for policy implications they are taken to be efficient. It is admitted that the model is a Utopia but since it cannot be falsified it is taken to be real.
Traditionally, science aimed to ‘save the phenomena’: a theory should be able to explain accepted observations or facts. In economics the principle is ‘save the theory’. If things claimed by the theory cannot be observed, occult, unobservable entities are posited by which they exist anyway.