Yes. An utter and total fraud.
First of all, it has no predictive power. You can predict what somebody will do if you assume that he is rational. Yes, you will have to make allowances for that person's circumstances, including his intelligence-level and personal disposition. But what you'll find is that people will tend to align their personalities and intellects with the dictates of rationality.
Yes, some are smarter than others, and even the smart ones make ghastly mistakes. But the mistakes are just that--mistakes. Whenever someone screws up, you can read all kinds of biases into it. But the litmus test is that if you try to predict what people will do on the basis of these same biases, you tend to get nowhere.
Pick any one of the biases that behavioral economists talk about, e.g. the 'anchoring' bias, the 'adjustment' bias, etc. Yes, people have those 'biases' in some contexts. But when circumstances make it clear that it is in one's interest to drop a given bias, people do so right away. Even the most inveterate racist or sexist will work with anyone of any race or gender if there is unambiguous evidence that it is profitable to do so, and the same holds of the 'biases' that behavioral economists talk about.
There is also a difference between a 'bias', as behavioral economists use this term, and entrenched irrationality. Suppose that you stripped someone's mind of all the 'biases' that behavioral economists talk about--the 'familiarity bias', the 'proximity bias', etc. That person would have to evaluate every situation de novo, which wouldn't be computationally feasible and would leave you in a state of a complete cognitive paralysis.
Contrariwise, when people have to get rid of those biases, they do, and they tend to do so quickly. If people were paid to do well on the various rationality test administered by the likes of Kahane and Thaler, they'd get very good at those test very fast.
And what about those tests? They weren't really tests, so much as they were pitiful attempts to find proof of biases already posited. For example, Kahane would ask people to multiply long strings of numbers, sometimes arranging those numbers from greatest to least, sometimes from least to greatest, sometimes with the even numbers at the beginning, sometimes with the odd numbers at the beginning etc. People were getting the right results. So Kahane made the strings longer and longer until people started making mistakes, and then he managed to find weak evidence of some kind of 'systematic bias.'
And if you use those 'systematic biases' to predict how people will multiply, you will tend to be wrong. And you'll tend to make the right predictions if you operate on the assumption that they multiply in a rational manner, give or take some insignificant deviations relating to fatigue, intelligence, and maybe a few 'irrational' biases of theirs.
If you consider any action on any person's part, you can identify a respect in which it is irrational. So and so reaches for the fork, instead of the spoon. They're 'irrational' since, at this point in the meal, they should having the soup not the salad, or whatever.
And that's what 'behavioral economists' are doing. They can't find robust, predictive 'biases', so they have to contrive biases. And then they get credit for being edgy, since the orthodoxy is that economic agents are rational. And of course when the market takes a turn, they can use it to prove the existence of these alleged biases.
But try predicting one of these downturns on the basis of these various 'biases.' You'll get nowhere. Economic crashes are sometimes predicted. But when this happens, the basis of the prediction has to do the structure of the market, along with the supposition that people are acting rationally. It can be rational for any given individual to sell or buy some junk security, even though the buying and selling of those same securities will collapse the market. This happens frequently, and some people see it happening and predict what will happen. But the relevant point is that the basis of the prediction is that individuals who are causing this collapse are doing so by behaving rationally, albeit negligently.
The biases that Kahane and Thaler talk about go away whenever people have a good reason to drop them. Which means that they aren't actually biases. They're heuristics that people hold onto to avoid excessive computational labor and that they drop when it is rational to do so. Behavioral economics is based on contrived studies and has a far worse track record at making predictions than the man on the street. But gives 'policy makers' and other 'experts' to insinuate themselves into people's lives. Now, thanks to the 'behavioral economics', an 'expert' on rationality gets to decide how your school cafeteria is configured and what kinds of books kids should read. This 'expert' is saving people from their own prejudices, you see.
And there it is. This phony non-predictive discipline stigmatizes the entire population as stunted and prejudice-driven, giving slimy bureaucrats, like Cass Sunstein, a chance to slap labels on them and put them bureaucratic boxes. And that's what 'behavioral economics' is about. As a theory, it's weak. As a bureaucratic cudgel, it is highly effective.