Kevin Drum has a post discussing whether the Fed could achieve a 4% inflation target. The post discusses some of Brad DeLong's idea, and then concludes as follows:
So if I'm reading DeLong right, it's not clear that the Fed could engineer 4 percent inflation at all right now. Maybe Scott Sumner has a bright idea about how we could do this.
This is always a difficult question to answer. Not because the solution is difficult---it's incredibly easy---rather because it's difficult to explain the solution to others, as most people think about policy in terms of cause and effect, not needed effect and implied cause. The simplest solution is to commit to buy T-bonds (and, if needed, Treasury-backed MBSs) until TIPS spreads show 4% expected inflation. At that high an inflation rate you don't need much QE, because the public and banks don't want to hold much base money.
But most people don't see things that way. They want to know how the policy can work assuming the public expects it to fail. I don't see why the public would expect it to fail if the TIPS market expected it to succeed, but let's put that aside. The fear is that even if the Fed bought all $20 trillion in T-bonds and Treasury backed MBSs, we still wouldn't get 4% inflation. If they adopted a 4% inflation target, I actually don't think the Fed would need to buy even one dollar more in debt, just refrain from raising rates for a few more years and stop paying interest on reserves. But let's say I'm wrong. What then?
Then you put a negative 2.75% interest rate on bank excess reserves, including vault cash holdings above a certain minimum threshold. That dramatically reduces the amount of QE required, as all the base injections go out as cash held by the public (and of course some of that goes back into safety deposit boxes, which is not the same thing as vault cash.)
I very much doubt that this hypothetical would ever come into play, but the thought experiment shows the Fed hasn't even done 10% of what it could do, so no one should ever be pessimistic about the ability to hit a 4% inflation target based on actual central bank policies in recent years---they haven't even come close to testing that ability.
Of course anything is theoretically possible, and if I'm completely wrong then the government would face a choice, either undershoot the 4% target, or let the Fed buy things they currently are not allowed to buy. But I would strongly encourage Kevin Drum to rethink his pessimism on this issue. The passivity of real world central banks should not lead us to doubt what a really determined monopoly producer of intrinsically worthless fiat money can do, if sufficiently determined. They can debase their currency, if they wish to. As James Hamilton once said (not exact words) "If the Fed doesn't know how to create inflation, then put me in charge."
That's how I feel too. Fortunately there is no need to---the Fed insists it could do much more stimulus if it wanted to.