I once explained this to someone like this: Suppose you were to sit down at length with [insert favorite sports hero here, I'd pick Brett Favre just for fun]. He is probably one of the worlds experts on how to read a defense, what the correct throwing form should be, what exercises to do in the off season, what qualities to make for a good wide receiver, etc.
Now imagine you ask him "Who is going to win Texas vs. Alabama this year" Is he more qualified to answer that than pretty much anyone you will probably meet in your whole life; yes. Does all of his expertise in other areas help in making that decision; yes. Is the guy going to do statistically better than a coin toss? Probably not. We don't need to "rethink quaterback theory" or "throwout freshman textbooks" because of this though as people seem to be fond of claiming.
This is true, and a good point. One thing we do need, however, is to take the opinion of prominent economists with a little bit more salt. I remember when invoking the name Greenspan among those interested in economics was like invoking the Name of the Lord - there was no higher authority. Now good ol' Al is getting hung out to dry and, in my opinion, trying to shunt blame away from himself as much as possible.
There shouldn't be a cult of celebrity in any science, economics least of all. This is what the NYT author should have argued against rather than economics in general. Before Farve went to play for Minnesota, how many Packers fans do you think would have bet against his prediction in the Texas vs. Alabama game if it was a very public prediction? The line in Wisconsin would be outrageously off as bookies bumped it more and more away from Farve's prediction. Go to Vegas and it's probably a lot closer to the money. This decision distortion is a real problem that is wreaking real havoc on our system of exchange.
I think it makes sense that some economists are a lot better at their jobs than others. Many may fall into ideological ruts, because industry and government have strong interests in the outcome of this social "science".
Now imagine you ask him "Who is going to win Texas vs. Alabama this year" Is he more qualified to answer that than pretty much anyone you will probably meet in your whole life; yes. Does all of his expertise in other areas help in making that decision; yes. Is the guy going to do statistically better than a coin toss? Probably not. We don't need to "rethink quaterback theory" or "throwout freshman textbooks" because of this though as people seem to be fond of claiming.