Taylor Rules and NGDP Targets: Skeptical Davids by Thoma

Andy Harless comments

AFAICT David Altig gets it completely wrong. His argument that "nominal GDP targeting...provides a poor nominal anchor in an environment in which there is great uncertainty about the path of potential real GDP" is valid for NGDP growth rate targeting (and indeed is one of several convincing arguments against such a regime). But very few people are currently advocating a regime of NGDP growth rate targeting. The proposals coming from Sumner, Romer, Woodford, etc., etc., are all for NGDP level path targeting, which is a completely different ball game. Compared to the current system, NGDP level path targeting would provide a much more effective nominal anchor. 

In general, level path targeting provides a much more effective nominal anchor than growth rate targeting -- which is essentially what we have now, a system that targets the growth rate of the price level, otherwise known as the inflation rate. If one insists on using a growth rate target, there are any number of strong reasons to prefer the current system over an NGDP growth rate target. But once you accept using a level path target, those objections disappear, and I think there is a very strong case for preferring an NGDP level path target over a price level path target. I also think that, on balance, the credibility gained by using a level path target rather than a growth rate target is worth the cost in terms of occasional instability (when you have to overshoot deliberately in one direction to make up for an earlier undershoot, etc.). I didn't think that before 2008, but at this point I'm convinced.