You're mixing up two different notions of inflation here.
There is the old-fashioned definition of inflation which calls any expansion of the "money supply" inflation. (With "money supply" in scare quotes because it's a pretty arbitrary concept, considering the multitude of very different definitions that exist.)
Then there is the modern definition of inflation, which is simply a different word for aggregate changes in the price level.
Those two notions of inflation are only superficially linked, and arguably, the modern definition makes more sense because it is what people are really interested in outside of economics.
On the topic of NGDP targeting, the blog Unlearning Economics has two enlightening posts: http://unlearningeconomics.wordpress.com/2012/05/08/on-the-l... http://unlearningeconomics.wordpress.com/2012/05/12/more-on-...
If the central bank targets 2% price level inflation, the economy grows 2% (witch would mean 2% of deflation in prices under the productivity norm) the central bank then has to go on and print enough money that it in a non growing economy would have hit 4% nominal inflation.
I don't see where I mix up the concepts I am always talking about price inflation.
Edit: On the blogpost, I don't think the autor really understands what NGDP targeting is about. Its about the CB not distorting relative prices, its seeing whats going on in prices. With IT you mess up the function of the price system, Hayek teaches us that prices are information. The productivity norm is basically what would happen without a central bank.
You mean "actually produce 4% inflation in nominal terms"
Im very much against that, the productivity norm (what is discribed in the less then zero book) would acctually produce devlation most of the time. If you want it in the 'proper' terms it would mean just setting a constant MV and let Py jump around as much as it wants.
But generally I agree with out, the post is on to something. Specially about defaulting. It is better to default, start fresh. This way the people that badly invested get punished and you get the chance (see iceland) to implment a more modern govermnet (iceland got a new constitution).
There is natural and artefical inflation and deflation. You get for example natural inflation when you have supplyshocks (If somebody blow up intel CPUs would be more expensiv) you get artefical inflation when you print money. The first is just the market reacting to 'something'.
The same goes for deflation, you have natural (good) deflation when you become more productive (you can see this for example in times of the (real) gold standard when moneysupply was almost constant.
Bad deflation is when you activly contract the money supply (this is the great depression story and the reason why people think deflation is generally bad).
What can we conclude out of this?
The shock to the economy was pretty big, we shuld have seen natural inflation happening, this is however still represt since the ECB pushes that down by creatning bad deflation (in real terms). In nominal terms this means 2% inflation, in real it is deflation.
When central banks target inflation (say 2%) and the economy is growing around 2% (thus 2% deflation) they acctully produce 4% inflation in nominal terms. This is all works out relativly good and is not to harmful as long as the real terms are not to big. If the become big (say a economy that is shrinking 10% or more) targeting inflation can be very harmful.
The price level should reflect the economy and not what some guy think is the right growth of the price level.
The idea would be to target NGDP or in terms of the formula MV=Py, it would mean holding MV constant (or let it grow 3-7% depening on your opinion on some other matters).
The school that supports this would be: http://en.wikipedia.org/wiki/Market_monetarism
This would mean that the ECB should look at the grow/contraction of the economy and setting the price level accordingly. That would mean in the current case that they would have do QE to hit the 'right' pricelevel.
I would recomend this book: Less Than Zero: The Case for a Falling Price Level in a Growing Economy http://mises.org/books/less_than_zero_selgin.pdf
Edit: I just found a nice post that shows a nice example of a case in history when central banks tried to recontract after the had inflated befor, this is in many ways simular with what we have today:
Danish and Norwegian monetary policy failure in 1920s – lessons for today
http://marketmonetarist.com/2012/06/12/danish-and-norwegian-...
Edit: Correction of a error point out in comments.