Note that in his latest column, he says that breaking up Microsoft is not economically optimal, on the basis of 'textbook economics'.
In the second-to-last column, he says:
Consider, on one side, really tough issues -- where there are
plausible arguments on both sides, where nobody really
knows how to measure the tradeoffs. Should Microsoft be
broken up and, if so, how?
So, which is it? Or should we take him literally, and assume that that *textbooks* specifically mention Microsoft, and the consequences of breaking it up, while the *economists* themselves are uncertain?
Overall, both columns are good, in that he points out that many are quick to invoke/deny economics, as it suits their needs of the moment.
However, Prof. Krugman himself is an excellent example of that syndrome (his column on the Microsoft break-up, and his column on rent control and SF).
Barry