Economics is a branch of moral philosophy and very consciously uses persuasive rhetoric to dictate policy. One of the tools used is hand-waving hidden inside differential equations to produce models that would be useless in any other discipline.

I don't think I've ever seen a mainstream economic prediction that was actually correct.

It's not hard to understand why. Reducing all the sectors of a complex economy to crayon-drawn measures like "inflation" and "unemployment" - which aren't even measured with any consistency - is like trying to predict the weather in the Bay Area using a single weather station for the entire continental US, which is conveniently located on the wall of a cow shed in Kansas.