The latest by Paul Krugman here: How Did Economists Get It So Wrong?
So what, pray tell, have we learned from the blowhard this time?
First thing: Krugman is not an economist. Evidence: Throughout his long rant, he is careful in referring to economists as "they." So whatever went wrong, please don't blame Krugman. On the other hand, not being an economist does not appear to place limits on his profound knowledge of how the profession should reorganize its thinking. Please show us the way, o wise one.
But first, what went wrong exactly? His claim is that the profession fell in love with its mathematically elegant models; and adopted a belief that unfettered markets achieve the best of all possible worlds (Dr. Pangloss). Evidently, this latter belief, supported by unsubstantiated modeling, translated into policy advice with predictable consequences.
Funny though: My reading of economic history suggests that economic crisis preceded mathematical modeling. Moreover, the phenomenon seems to transcend institutional regime (economic crises are endemic to "planned" societies as well). And as for how the philosophy of free market capitalism has manifested itself in reality--well, this is utterly laughable.
In any case, his claim that "Freshwater economists are, essentially, neoclassical purists" is so far off the mark as to make one question his academic credentials. The neoclassical framework is certainly viewed as a benchmark; but almost all serious work that I am aware of regularly departs from its basic tenets (in particular, by explicitly modeling the problems that arise when commitment is limited and information is private). There is, in fact, much work being done in taking institutions seriously, in modeling environments where trade is subject to search frictions, and in identifying potentially beneficial policy interventions. Krugman would not be aware of this, of course, as he spends all his time writing op-ed pieces for the NYT instead of actually engaging in difficult research questions.
It is true, however, that most of the profession adopts the view that individuals are "rational;" at least, in the sense that we model people trying to the best they can (according to a well-defined objective) subject to the constraints imposed upon them by the economic or institutional environment. I believe that this assumption is employed for three reasons:  the idea that people try to do the best they can subject to limitations does not sound crazy;  the hypothesis admits all sorts of "crazy" equilibrium behavior anyway; and  it is hard to know what the hypothesis should be replaced with. In particular, while there is only one way to be "rational," there are an infinitely many different ways to be "irrational."
To give you just two quick examples, Krugman offers his Capitol Hill Babysitting Cooperative anecdote as some sort of puzzle for "neoclassical" economists. I doubt, however, whether he has read this. Or, if you believe his rant the mathematically inclined are oblivious to possibility of economic catastrophe; read this.
As an alternative, Krugman proposes the methodology of behavioral finance. Basically, the approach there is to simply assume that people behave according to some (theoretically imposed or empirically estimated) rule of thumb. In less polite language, assume that people are "stupid." Personally, I believe there may be much to be learned by this approach; and I welcome the fact that a part of the profession devotes some time exploring its implications. But one gets the sense that Krugman prefers this approach because by adopting it, we admit that the general population is stupid and is therefore in need of guidance. This guidance, quite naturally, is to come from self-appointed philosopher kings, like Krugman (consulting $).
On one point, Krugman is correct: There was a growing complacency among many in the profession (and elsewhere). He is incorrect, however, in suggesting that this complacency was the product of economists falling in love with their mathematical models. The majority of the profession continues to whittle away at difficult problems in relative obscurity. The complacency, in my view, stemmed from people like Krugman -- people who poo-poo those of the profession engaged in exploring difficult problems in a rigorous manner.
Krugman prefers his simple equations--an IS curve, and LM curve, and all his "fudge factors" to explain the world. Not much else is needed when you know that the world is stupid; and that you alone hold the answers.
Unfortunately, people are evidently too stupid even to recognize Krugman's genius (confirming his hypothesis in his own mind, no doubt). Is this the source of his thinly-disguised bitterness?