I agree with his diagnosis. But his prescription, to return to some sort of old-time Keynesian religion, is too vague. It would be precise if one could speak of the old-time Keynesian religion, but unfortunately schisms erupted from the very beginning in the 1930's and have continued to this day. Economists are no more ready to agree on what is Keynesian than Iraqis are ready to agree on what is the true Islam.
It is odd that Krugman wrote this for the New York Times. It belongs in the new American Economic Association journal Macroeconomics. An imaginative editor would open up the pages of such a journal to heterodox points of view. But, come to think of it, that journal is edited by Blanchard. Never mind.
Krugman trots out his baby-sitting co-op story, which is roughly equivalent to my hydraulic macro fable (he wrote his fable before I wrote mine). However, Krugman does what I suggest not to do, which is make a big deal out of the problem of money as a store of value (the baby-sitting coupons on his story).
The odd thing about the baby-sitting co-op is that there is no possibility of investment. There is nothing you can do today that will increase production of baby-sitting services tomorrow. Because there is no investment, there is no possible place for savings to go. So saving is contractionary. But why should that pathological situation carry over into an economy in which there are real capital investments available?
In the real world, if you want to have consumption for next year, there are things that are produced today that can give it to you. Buying a car can give you transportation services for next year. Buying frozen food can give you food for next year. Buying a house can give you shelter next year. Investing in securities can provide funds to firms that purchase capital equipment that will yield consumable goods next year.
What I propose is that instead of thinking that saving exceeds investment in some aggregate sense, we consider that the market is in the midst of a great recalculation, due to the fact that suddenly there is a misalignment between production plans and consumption plans. Two years ago, builders were expecting consumers to absorb 2 million new homes a year. Today, they know that's not happening. Consumers would like to increase future consumption of something other than housing services. The recalculation problem starts with the fact that nobody is quite sure what consumers would like this future consumption to be. Health care services? Fancier smart phones? Devices enabled by nanotech or biotech that has not yet been perfected?
Along the way, all sorts of changes in wages and prices will take place. These will send signals that will lead people away from some occupations and toward others. But the process is going to take years, and meanwhile there will be idle resources. Government might be smart enough to employ the idle resources for a while and then release them when the market is ready. But I don't think that Congress designed the stimulus package with that in mind.