It has no predictive value and therefore no explanatory value.
And the reason it has no predictive value is that its entire purpose is to validate an assumption that is false, that assumption being that people are not rational economic actors.
The truth is that people are rational economic actors.
People drop out of college if they stand to gain by doing so, and they stay in college otherwise.
People switch brands of toothpaste if they stand to gain by doing so, and they stick with their current brand otherwise.
There are meaningful deviations between actual and rational economic conduct only in connection with highly artificial contexts (e.g. the buying and selling of securities) and even then only relative to inappropriate yardsticks (e.g. portfolio-performance after a period of one year, as opposed to twenty years).
The so-called prejudices cited by the likes of Kahane and Thaler are held onto only as long as they are useful, showing that they are not prejudices.
When people behave in ways that are rational but not to the liking of policy-makers, behavioral economics provides said policy-makers with superficially plausible proofs of the irrationality of said conduct.
And that is the discipline’s sole purpose.
Behavioral economics could be a legitimate discipline.
But it isn’t.
It isn’t, because, in its current form, it is a political cudgel.
And that’s all it is.