But he doesn't really buy my bizarre monetary theory. He writes, "I see the Fed as controlling nominal gdp but not always real gdp."
Let me start this way. There is a short run in which monetary policy cannot control nominal GDP, and there is a long run in which monetary policy cannot control real GDP. That is, there is a short run in which M affects only V and a long run in which M affects only P.
Suppose we measured GDP on a weekly basis. What can the Fed do this week that would affect nominal GDP next week? I say "nothing." Maybe Scott Sumner wants to say, "the Fed can change expectations," in which case we can start our argument there. I don't think that the expectations that matter for next week's nominal GDP are expectations that can be changed quickly. Or maybe Sumner will concede that next week's nominal GDP is given, so that for next week any big increase in M will result arithmetically in a big decrease in V.
In the long run, the neutrality of money says that M affects P, not Y. We can go with that long-run neutrality, although V changes a lot over time due to new methods for processing transactions.
So, the question is whether there is a medium run in which M affects Y. My bizarre monetary hypothesis is that the answer is "no." That is, I believe that the medium run usually looks like the short run, in which changes in M show up as changes in V. The medium run only looks different if the central bank is engaging in a regime shift, changing the long-term trend of M and P.
The long-run trend of M affects the long-run trend of P. If people get used to low inflation, there will be low inflation. If people get used to high inflation, there will be high inflation. It takes a long time to change what people are used to.
In my hypothesis, the central bank can only change behavior by undertaking a long term regime change. This necessarily involves a long lag, since people only internalize regime changes by observing the outcomes over a period of years. It is unlikely that the central bank can fight a short-term cyclical downturn with a regime change that requires several years to take hold.
In short, the bizarre monetary hypothesis that I am proposing is that monetary policy only affects nominal expenditure in the long run, and in the long run monetary policy affects inflation rather than real output.