What is the Fed Thinking, 2?

"So, the Fed could be thinking that setting an inflation dial to 3 percent is not an option. For small changes in Fed policy, the inflation rate will not be affected. And if the Fed goes beserk [sic], the result will be to move us into the regime of high and variable inflation rates, which will actually be worse."

Wouldn't level targeting be a credible way to avoid this problem? (Though I'm not sure what the unintended consequences of level targeting would be.)

Oh I think that is quite plausible. Also, in our current economic environment I think targetting inflation is even more difficult than pre 2008. There is just so much going on and so many drivers have rapidly changed course in the last year or 2. To the extent that certain inflation drivers (other than Fed policy) kick in with some sort of time lag, and I think some do have a significant lag, trying to forecast inflation out a few years is unusually challenging.

There's also a time-lag problem of a different sort I think. Some of the factors imbedded in the economy and monetary policy point towards short-term disinflation or maybe even deflation. Other more longer-term indicators seem to me at least to point towards inflation. Trying to reconcile the 2 is messy and potentially dangerous.

I can buy this theory as long as people are just flying blind without any guidance from the Fed about what its intentions are, or if they don't trust the Fed. But the Fed can state its intentions, and given the experience of the past 30 years, people have reason to trust it with respect to inflation policy. The Fed has developed plenty of credibility, and it could put that credibility to good use by instituting a policy that commits it to high inflation in the intermediate run but low inflation in the long run.

That's not as hard as it sounds, because all the Fed has to do is commit to a set of price level or (more effectively, under the circumstances) nominal GDP targets that continues the trend that was in place before the downturn began. For the first few years, the Fed would probably not be able even to come close to those targets. Indeed the gap between the actual price level or NGDP and the target would probably increase for the first few years, which would result in the Fed's being committed to a higher and higher inflation (or NGDP growth) rate in the subsequent years. Eventually, investors and businesses would catch on and start expecting more inflation and investing in real assets.

Then, as higher inflation starts to happen and the path of the actual price level (or NGDP) starts to converge toward the path of the target, the Fed's implied inflation (or NGDP growth) target would automatically go down. It wouldn't be perfect: yes, the Fed would eventually probably have to induce another recession in order to slow inflation again once it rises. But the recession probably wouldn't be a severe one given that its price implications would be heavily anticipated.