That is the discussion among the banking economists at the moment. Kling, Sumner, etc. In short, which variable should the Fed watch when setting monetary policy, Nominal (non-inflation adjusted) GDP, real short term interest rates, or the money base in circulation. I propose here that the discussion resolves around the response time of the Fed, not the target.
We are really talking about the TIPs aberration in the chart above. As I pointed out, this aberration is the result of the economy changing faster than the adaption period of the Fed. There is a six month bandwidth limit in the financial system, it cannot be avoided regardless of the targeting regime.
Even George Selgin's competitive monetary system would only smooth out the pulse, but the delay remains.
Why do we have a six month noise limit in aggregate measurements? Because the cost of adding more bandwidth is enormous. How enormous? Look at the cost of our central bank in dealing with the crisis in late 2008. The central bank and Treasury were in the headlines many days, high level visible meetings all day among officials in Washington. We simply cannot afford the cost of collecting aggregate information with timelines less than six months.
Information theory gives us a clue. If finance operates with 5 bits of precision, and suddenly needs 6 bits, then the number of transactions the finance needs goes from 5Log(5) to 6log(6). As long as the economy is operating within a stable "corridor" we will want the Fed operating with low precision, and we are willing to suffer the consequences when change happens too fast.
No comments:
Post a Comment