The past few weeks have settled, to my satisfaction at least, a long-running debate on this very topic. Rather than targeting inflation, central banks should keep nominal incomes growing on a pre-announced path: say 5 per cent a year.
Pointer from Scott Sumner, who of course has advocated something along these lines for quite some time. My thoughts:
1. According to textbook macro, inflation can always accomplish the same thing as fiscal stimulus. In fact, there is no macroeconomic model which says both that deficit spending is expansionary and higher inflation is not. This is a point of economic theory that is not understood by the public.
2. The central bank can always cause inflation if it wants to.
Point (2) should be uncontroversial. However, the idea of a "liquidity trap," if it means anything, means that the central bank cannot cause inflation if it wants to. When someone says that we are in a liquidity trap or the Fed is out of ammunition, they should be obliged to defend the proposition that the central bank can no longer cause inflation, even if it wants to.
Pundits who advocate for fiscal stimulus rely on the public's ignorance of point (1) to make their case with laymen. To make their case with economists, they rely on creating fear and doubt about point (2). They don't convince anyone with logic, but sometimes sheer repetition and contemptuous bullying do the trick.
Together, points (1) and (2) imply that fiscal stimulus is not necessary in order to increase aggregate demand.
3. It could be that aggregate demand is not the main factor in our current troubles. See my posts on PSST.
4. It also could be that inflation cannot be fine tuned. That is, when inflation starts to rise, the public takes steps to economize on the use of money, which causes velocity to rise. Thus, there are two regimes--one with low and steady inflation; and one with high and volatile inflation. An inflation rate in the range of, say, 4 to 8 percent, is not stable. Once you get above 4 percent, you either have to squeeze inflation out altogether or watch it rise toward double digits. I have argued for this two-regime story in the past, and I continue to think it has some validity.
If (3) and (4) are correct, then monetary expansion may have adverse results. It will not increase employment by much, and it may push us into the high-inflation regime.
All of this said, I would like to see monetary expansion tried. The probability that (3) is correct is not 100 percent, and the probability that the textbook model of AS-AD is correct is not zero. Given the severity of the unemployment situation, monetary expansion is worth a try, in my opinion.
The simplest approach to monetary expansion would be for the Fed to stop paying interest on reserves. However, in addition, the Fed could purchase more assets.