From Harvey Mansfield, political theorist, and it could be subtitled "How then should we live?" (the actual subtitle is "Is the overly predicted life worth living?"). Via Tyler Cowen, who linked to the article but did not venture a response. Here's Mansfield's conclusion:
Economics needs to stop trying to duck responsibility for what it recommends. It needs to examine the whole of life and to focus on the virtue or virtues of different ways of life. It should give over talk about "preferences," as if human desires were given facts unaffected by the science of economics. It should abandon the crude positivism that claims that one can study facts without giving advice, or that one can confidently predict without causing people to believe in one's predictions. It needs to replace its false modesty with true moderation.
I doubt Mansfield will receive many answers from economists. All through my economics education I was constantly told that the study of economics is about what is possible to do, not what should be done; that question was reserved for the theorists, philosophers, and clerics. Modern economics is too tied to positivism to take Mansfield's advice, even if it wanted to (and it almost certainly does not). The old political economists -- J. S. Mill, Smith, Ricardo, Marx -- were centrally preoccupied with just this sort of normative question, but modern economists are proud to leave them to others. There are still a few strands of political economy happy to engage the debate, but all too often conclusions are a pure function of priors, so the interaction leaves something to be desired.
I'm not as sure as Mansfield that this state of things is so terrible. After all, there's something to be said for gains from specialization, and different scholars have different comparative advantages. Expecting quantitative methodologists or formal theorists to spend much time parsing normative concerns is just as silly as expecting philosophers to learn advanced game theory. We can have both: positivists to tell us the range of possible outcomes, and normativists to help us decide which outcomes to pursue. All to say, I don't think Mansfield's intended audience is capable of responding the way he wishes they could, and that's probably okay.