Tate Lacey of the Cato Institute has an interesting post discussing Jerome Powell’s recent Jackson Hole speech on monetary policy:

Powell constructed his remarks around a nautical metaphor of “shifting stars.” In macroeconomic equations a variable has a star superscript (*) on it to indicate it is a fundamental structural feature of the economy. In Powell’s words, these starred values in conventional economic models are the “normal, or “natural,” or “desired” values (e.g. u* for the natural rate of unemployment, r* for the neutral rate of interest, and π* for the optimal inflation rate). In these models the actual data are supposed to fluctuate around these stars. However, the models require estimates for many star values (the exception being desired inflation, which the Fed has chosen to be a 2% annual rate) because they cannot be directly observed, and therefore must be inferred.

Powell is worried that monetary policy needs to “steer” using various stars that are not clearly visible.

In addition:

Two years later, Powell said that traditional rules were backward looking, but that monetary policy needs to be forward looking and not overly reliant on past data. Upon becoming Fed Chair early this year, Powell made it a point to tell Congress he found monetary policy rules helpful—a sentiment he reiterated when testifying on the Hill last month.

The good news is that there is a monetary policy rule that is forward looking, not concerned with estimating the “stars,” and robust against an inflation fixation. I am referring to a nominal GDP level target, of course; a monetary policy rule that has been gaining advocates.

Lacey has identified a number of areas where Powell seems to (at least implicitly) understand the advantages of NGDP targeting.

1. One well-known advantage of NGDP targeting is that inflation targeting does not do well during periods of supply shocks.  Inflation proponents try to remedy this problem by adding an estimate of “output gaps” to their policy approach, as in the famous Taylor Rule.  But output gaps are notoriously difficult to estimate; indeed the Fed recently assumed that the natural rate of unemployment was well over 5%, a view that now seems incorrect.  Under NGDP targeting, there is no need to estimate output gaps.

2.  It is possible to construct a modified Taylor Rule for NGDP targeting, in which case you’d still need to estimate the natural rate of interest.  However, Powell correctly notes that any Taylor Rule-type regime is excessively backwards looking, and policy is better off being forward looking.  This leads to Lars Svensson’s call for “targeting the forecast”.  If you target the market forecast of NGDP then there is no need to estimate the natural rate of interest.  BTW, Alex Tabarrok recently discussed some very interesting research, which indirectly supports Robin Hanson’s claim that policymaking is most efficient when guided by prediction markets.

Of course NGDP targeting has many other advantages as well:

3.  It does a better job stabilizing labor markets than does inflation targeting.

4.  It does a better job stabilizing financial markets than does inflation targeting.

5.  NGDP level targeting is a more rules based approach to policy than “let bygones be bygones” inflation targeting.  Ironically, if the BOJ were to adopt a NGDP level target tomorrow (say along a 3% NGDP growth path), it would actually be easier to predict the Japanese price level in the year 2038 than it is under the current 2% inflation target.  That’s because when the BOJ misses the 2% inflation target (as they do frequently) there is no “make up” to get back on track.  Under 3% NGDP level targeting, I’d expect the Japanese price level to be about 40% higher in 2038.  Right now I have no idea whether it will be 40% higher or 4% higher.

6.  NGDP targeting is much easier to explain to the public.  In 2010, Bernanke told the public he was trying to raise inflation to 2%, and 99% of Americans had no idea why he was trying to raise their cost of living.  If he had instead said he was trying to raise the incomes of Americans by 5% to promote recovery, people would have understood the rationale for Fed policy.  When people think of inflation, they think of the supply-side inflation that makes Americans poorer, not the demand-side inflation that the Fed is actually trying to influence.

Jerome Powell seems to have excellent instincts.  My prediction is that the Fed will not immediately switch to NGDP targeting.  Rather when we are about to enter the next recession, the Fed will use NGDPLT as a temporary expedient to create more stabilizing expectations, citing the research of Michael Woodford and others.  When and if it is successful, they will quietly forget about the old inflation-targeting regime, and we’ll be in a new world of NGDP targeting.  And then I can finally retire and read all those novels piling up on my shelf . . . Thank God!

HT:  Alvin Rabushka