I see there is quite a bit of concern about the Bank of England's quantitative easing measures, but I don't think it quite hits the target (i.e. I'm concerned about something different from other commentators). Central banks engage in quantitative easing all the time - they just usually refer to it by the name "interest rate cuts". For the way in which many central banks cut interest rates is to offer banks more money (i.e. print money), so the price of money (the interest rate) falls. On other occasions they engage in "quantitative tightening" - i.e. interest rate rises.
The key interesting thing in the current situation does not relate to the principle of quantitative easing but, instead, to the form or method of it. For once interest rates reach zero, one cannot employ the same methods to increase the quantity of money. So if they are to engage in monetary easing, they have to use "unconventional monetary policy".
Obviously we have less experience in the precise effects of unconventional monetary policy, so there is a good debate to be had about the best way to do it. But whether to do it is a variant of the usual sort of monetary policy debate - are we in a situation in which we want to tighten or loosen monetary policy? It's pretty obvious to almost everyone that we want to loosen (money supply is collapsing; inflation is projected to go well under target; the real economy is tanking) - though there might be some who would point to the loosening there has already been and say it should be allowed to work before we try more.
There is, however, an important difficulty related to uncertainties about the effects of unconventional policy in a deflationary environment. As Mark Field correctly points out, in deflationary episodes it is almost never the case that the amount of money, per se, is inadequate. Instead, what happens in deflation is that the money stops circulating so rapidly in the economy (the "velocity of circulation" falls) - e.g. because the banking system stops functioning normally. And he is right in that what really matters is the multiple of these two - the total amount of money circulated, usually called "nominal GDP".
Now, suppose that I care about the multiple of two things. Let's say, I care about the number of apples I have in my lorry times the price of each apple, i.e. in the total value of my lorry-load of apples. Then, if the amount of one of these things falls I can make up for that by increasing the amount of the other. For example, if the price per apple falls, I could keep the total value of my lorry-load of apples the same by increasing the number of apples I carry.
In the same way, if we care about nominal GDP (the amount of money times its velocity of circulation), then if the velocity of circulation falls we can make up for that by increasing the amount of money. That is what is proposed here: we make up for the deflationary pressure created by a fall in the velocity of circulation of money by printing extra money.
But there is a problem. For the velocity of circulation will, eventually, return to something not too far from its previous level. Although in the short-term velocity can be extremely sensitive to expectations factors, in the medium-term it must depend on things like the state of technology (say, whether we have cash-points), institutional factors (how the banks interact), and other "real" factors. A significant financial sector episode like the current one might well mean that the velocity of circulation will fall even in the medium-term, but it won't fall by anything like as much as it falls in the short-term (driven by negative sentiment).
Now go back to the apples. If the price of apples is low as I arrive with my lorry, then I will need extra apples to have a load the same value. But if I arrive with the extra apples, then the price goes up before I've sold them all, my extra-apples lorry-load will eventually end up being worth more than my original fewer-apples one.
In the same way, once the velocity of circulation rises back up to its medium-term level, the fact that we have printed extra money will mean that nominal GDP will shoot up, creating inflationary pressures. This effect is likely to be rather large and difficult to control.
Consequently, on the "exit path" - i.e. as velocity returns to its more normal medium-term level after the deflationary problems are over and the banking system starts to function more normally - one would expect to have quite a bit of inflation. This is sometimes called the "ketchup-in-the-bottle effect" - you keep bashing on the bottle by printing more and more money without obvious effect on prices, then you suddenly get a splurge when your bashing finally works.
To manage this, one needs a monetary framework that tells you what to do on the exit path from a deflation. Inflation targeting does not. A price-level target does. That is precisely the main reason I have argued previously that price-level targets were better at dealing with liquidity traps (i.e. zero interest rate or banking sector impairment scenarios).
