How Would Monetary Stimulus Help?

Edward Harrison and I have been cordially discussing what, if anything, the Fed could do to improve economic conditions.  Like many, he is understandably skeptical about the efficacy of monetary policy. He notes that households are repairing their balances sheets and the massive deleveraging they are doing along the way is putting a strain on the economy.  He wonders how the Fed could help in such a situation.   I, on the other hand, am more optimistic about monetary policy and believe the delevaraging is not much of a  constraint for the Fed.  Monetary stimulus was able to spur a robust recovery during the depths of the Great Depression from 1933 and 1936 and did so despite households delevaraging then too.  The question, then, is how could the Fed actually make a difference?  Below is part of an email exchange we had where I attempted to answer this question.  Harrison posted it on his blog and I thought it was worth reposting here.  Though I explain in the note how price level or nominal GDP level targeting could bring about a robust recovery in nominal spending, I really prefer nominal GDP level targeting for reasons laid out here. 

The key for the Fed to be effective is for it to radically change or "shock" expectations about future nominal spending.  The Fed would do this by announcing an explicit NGDP level (or price level) target where it publicly committed to buying up as many assets as it needed to return NGDP  (or the price level) to some pre-crisis level trend.  Since this would mean higher-than-normal NGDP (or price level) growth for some time (until the pre-crisis trend is restored), it would imply the Fed would be purchasing a lot more assets.  This would "shock" households and firms into spending their money assets on other riskier assets (like stocks and physical capital) as well as goods and services lest their money assets lose value from the higher expected inflation during the catch up growth period.

Some points:

1. Even if only the riskier financial assets are initially purchased, the sellers of the financial assets would now have money balances they will want to unload because of the higher inflation.  They too will start buying up other assets, goods, and services.  Eventually, then, the money assets will buy capital, goods, and services. This requires, though, that the Fed continues to buy up assets--including longer-term treasury securities-- until this happens, which is the very point of having a spending target like a NGDP level target--keep adding stimulus until nominal spending hits its target.
2. The key is to have a level target, not a growth rate target.  A growth rate target like inflation targeting lets bygones be bygones.  A level target allows for catchup growth and vice versa when there is a deviation from the target.  This allows for more aggressive monetary action in the short term, but at the same time creates more certainty over the long term since the policy aims for a target level growth path.
3. Having an explicit, well communicated nominal target means the Fed does not need to commit up front to an explicit  dollar amount of asset purchases.  This makes the policy less controversial, eliminates the need for successive rounds of ad-hoc QEs, and if credible will cause the markets to do the heavy lifting (i.e. if the markets believe the Fed is serious about higher nominal spending and inflation, they will respond on their own before the Fed purchases assets and thus lessen the needed amount of asset purchases by the Fed).
4. Not only have corporations, but households too have  built up relatively large share of liquid, money-like assets since the crisis.  All the talk about household deleveraging ignores that they have also been rapidly building up the share of liquid assets and thus curtailed spending.  The process above should help reverse this development.
5. Once the nominal spending takes off, there will be real economic gains, not just inflation, given the amount of excess economic capacity. (i.e. the increased spending will put unemployed resources back to work).
6. Because of (5), expectations of higher real growth will further reinforce current nominal spending and increase the demand for loans from banks.  Banks will respond to the increased demand for credit by extending loans.  The money supply grows in turn.