Now that I’ve actually read the full Hatzius/Stehn Goldman Sachs case for an NGDP level target and additional QE, I can see what they’ve added to informal discussion of these ideas from Scott Sumner, David Beckworth, and others, namely a quantitative estimate of the impact. They put together a “toy model” with the results you see below:
One interesting consequence of this model that may appeal to some Fed folks is that they say the more rapid recovery means that the Fed would exit the zero interest rate policy faster under the NGDP/QE scenario than under the baseline scenario. Basically if you’re uncomfortable with the idea of a prolonged period of “easy money” (i.e., low interest rates) you should push for aggressive stimulus to get us back to full employment.
But you ask, what is NGDP level targeting? Basically it means the Fed would say “we would to get the level of total spending and income in the USA back to its pre-recession trend and we’re prepared to do a bunch of QE to get us there.” Some of the increase in total spending would presumably come from higher prices (as the economy revives, gas prices will go up and so will hotel rates and various other things) and some would come from more real output. The NGDP target is indifferent as to the exact balance. There are a bunch of reasons this is clever, but here’s one way to think about it. If you say to the world “damnit, we’re going to get more spending even if it’s all inflation” then even pessimists who think it’s Kenyan socialism rather than low demand that’s causing the recession will want to unload some of their cash. That very inflation panic will increase demand. But since the pessimists are wrong about the economy, the increased demand will mostly result in more real output and higher real incomes rather than inflation.
NGDP targeting is also politically easier than inflation targeting. You say “higher overall spending and income” rather than “higher prices.”