Let the Fed target a Quasi-Real PCE Price Index (QRPCE)
The Federal Reserve on Wednesday said it would target a long-run inflation target of 2%. Some of my blogging Market Monetarist friends are not too happy about this – See Scott Sumner and Marcus Nunes. But I have an idea that might bring the Fed very close to the Market Monetarist position without having to go back on the comments from Wednesday. We know that the Fed’s favourite price index is the deflator for Private Consumption Expenditure (PCE) for and the Fed tends to adjust this for supply shocks by referring to “core PCE”. Market Monetarists of course would welcome that the Fed would actually targeting something it can influence directly and not react to positive and negative supply shocks. This is kind of the idea behind NGDP level targeting (as well as George Selgin’s Productivity Norm). Instead of using the core PCE I think the Fed should decomposed the PCE deflator between demand inflation and supply by using a Quasi Real Price Index. I have spelled out how to do this in an earlier post. In my earlier post I show that demand inflation (pd) can be calculated in the following way: (1) Pd=n-yp Where n is nominal GDP growth and yp is trend growth in real GDP. Private Consumption Expenditure growth and NGDP growth is extremely highly correlated over time and the amplitude in PCE and NGDP growth is nearly exactly the same. Therefore, we can easily calculate Pd from PCE: (2) Pd=pce-yp Where pce is the growth rate in PCE. An advantage of using PCE rather than NGDP is that the PCE numbers are monthly rather than quarterly which is the case for NGDP. Of course the Fed is taking about the “long-run”. To Market Monetarists that would mean that the Fed should target the level rather growth of the index. Hence, we really want to go back to a Price Index. If we write (2) in levels rather than in growth rates we basically get the following: (3) QRPCE=PCE/RGDP* Where QRPCE is what we could term a Quasi-Real PCE Price Index, PCE is the nominal level of Private Consumption Expenditure and RGDP* is the long-term trend in real GDP. Below I show a graph for QRPCE assuming 3% RGDP in the long-run. The scale is natural logarithm. I have compared the QRPCE with a 2% trend starting the 2000. The starting point is rather arbitrary, but nonetheless shows that Fed policy ensured that QRPCE grew around a 2% growth path in the half of the decade and then from 2004-5 monetary policy became too easy to ensure this target. However, from 2008 QRPCE dropped sharply below the 2% growth path and is presently around 9% below the "target". So if the Fed really wants to use a price index based on Private Consumption Expenditure it should use a Quasi-Real Price Index rather than a “core” measure and it should of course state that long-run inflation of 2% means that this target is symmetrical which means that it will be targeting the level for the price index rather the year-on-year growth rate of the index. This would effectively mean that the Fed would be targeting a NGDP growth path around 5% but it would be packaged as price level targeting that ensures 2% inflation in the long run. Maybe Fed chairman Bernanke could be convince that QRPCE is actually the index to look at rather than PCE core? Packaging actually do matter in politics - and maybe that is also the case for monetary policy.