Greg Ip (once one of our Carleton students, I'm proud to say) is not opposed to NGDP level path targeting. But he is not very enthusiastic about how well it would work.
Greg starts with a US anecdote, about a businessman who told Paul Volcker that he didn't think Volcker would succeed in driving inflation lower, and had just signed a 3 year agreement to raise wages by 13% per year.
I'm going to start with a Canadian anecdote, that I heard from Chuck Freedman, previously Deputy Governor at the Bank of Canada. Inflation targeting was not something dreamed up by the Bank of Canada, or any academic economist. The idea came from the private sector. Business people told Judith Maxwell, Chair of the Economic Council of Canada, that it would really help them set wages if the Bank of Canada told them what inflation rate it was aiming for. Because they really didn't know. Judith Maxwell told the Department of Finance, which told the Bank of Canada, and so the Bank of Canada figured out a new policy, and announced an inflation target. All the academic stuff came later.
Right now, Americans must have some expectations about what the Fed is trying to do. But those expectations must be all over the map. Even "cheap talk" by the Fed could help coordinate those expectations into a common belief. The coxswain on a Humean boat can help keep the oarsmen in time, and at a time he wants, even if he lacks a whip. Because each oarsman wants to row in synch with the others. American oarsmen might all want to raise the pace, but no individual oarsmen will raise his pace unless he expects the others to do so too. A coxswain calling out a faster time lets them all raise the pace.
Greg asks how it would work:
"Second, a nominal target should encourage firms and workers to behave in a way that makes the target self-fulfilling. This channel is well established for inflation targeting: if workers and firms believe the Fed will keep inflation at 2%, they will tend to set prices and wages accordingly. Exactly how an NGDP target would be self-fulfillling is unclear to me: I haven’t seen good empirical or theoretical evidence linking NGDP targeting to the behavior of private actors."
The theoretical evidence is clear enough. Suppose NGDP has been growing more slowly that the target growth path, and so is currently below the Fed's target. If people expect future NGDP to increase to the target path and close the ever-widening gap, they must expect one of two things: either faster inflation; or faster real growth. Or, more likely, some mix of the two.
The inflation channel under NGDP targeting is basically the same as the inflation channel under inflation targeting. Higher expected inflation causes higher actual inflation and/or real output growth. The main channels are through the expectations-augmented Phillips Curve, plus the gap between nominal and real interest rates (for Keynesians), or the effect of expected inflation on the demand for money (for monetarists). There's really nothing new here, except: NGDP level-path targeting is more akin to price level path targeting. The longer the actual inflation rate is below the target inflation rate, the bigger the gap between the actual price level and the target price level, and so the stronger this expectations channel will be.
We have 20 years of empirical evidence showing how inflation targeting works to stabilise expected and actual inflation.
It's the second channel, via expected higher real growth, that is new. At least, it's not there in inflation targeting. But it's not really new. An increase in expected future real income increases current consumption demand. An increase in expected future demand for real output increases current investment demand, because firms will have an easier time selling the extra output produced in future by current investment. An increase in consumption and investment demand raises current output and/or prices. These are two very old ideas in economics. You could model them with an Euler equation IS curve, but you don't really need to. Old Keynesians 50 years ago would know what I'm talking about.
But perhaps what's really new about NGDP targeting, relative to the inflation targeting we're all familiar with (OK, that most advanced countries in the world are familiar with), is the mix between those two channels.
Let's face it; we don't know as much about the Phillips Curve as we wish we did. Easier monetary policy will either increase the price level or increase real output or do some mix of the two. In the short run, we think it will do a mix of both. But we don't know what that mix is. And when Greg reminds us of the Volcker tightening, he reminds us that we don't know what that short run mix will be. In the long run theory tells us it's 100% prices. In the very short run the world tells us it's mostly real output. But we don't really know the precise mix.
NGDP targeting is like targeting a 50-50 mix of the two. We still get the long run nominal anchor that theory requires. But in the very short run, the 50-50 mix is almost certainly more plausible and therefore more credible than the 100-0 mix that an inflation target calls for. Is 50-50 the exactly right mix? Almost certainly not. But it's simple, and acts as a natural focal point, which is what you need if you are trying to coordinate expectations. It's like your income, measured in dollars, only for the whole economy. And it's much more credible for the Fed to claim that it can hit some average of prices and real output than for it to claim it can hit just one or just the other. Opinions on the slope of the Phillips Curve really do differ, and rightly so in the short run.
So, if credible, it works. But will it be credible? Here's Greg again:
"Since 2007 the Fed has worked overtime to push employment higher and keep inflation from falling, which it can justify thanks to its 1970s vintage mandate of full employment and stable prices. Why would swapping its old framework for an NGDP target change this? Advocates claim this would justify a far more aggressive policy of quantitative easing. I am all in favor of more QE, but the Fed does not need a new framework to do that; its current mandate provides all the justification it needs.
So why doesn’t it? The Fed now finds itself in the odd position of being blasted from one side for doing too much and the other for doing too little. There is far more substance to the latter arguments than the former, but NGDP advocates base their arguments on a flawed premise: that with a different framework the Fed would have been less concerned about inflation and more about output, and would have thus eased more aggressively."
What exactly is the Fed's "old [existing] framework"? Professional Fed watchers don't know what the Fed is trying to do. How could we possibly expect ordinary people to know what the Fed is trying to do? Reading US commentary, both professional and amateur, shows that beliefs about what the Fed is trying to do are all over the map. Nobody can hear the coxswain's beat, and they all think he's beating totally different times. The oarsmen are all rowing at different speeds. "Structural" problems are a symptom, not a cause, of this coordination failure.
Within limits, it matters less what beat the coxswain calls, than that he does call a beat. Some at the Fed would prefer a slightly faster beat than others. But it would be much better to resolve that argument, and agree on a beat and stick to it, even if the compromise is not one's ideal. The average of two different coxswains would be better than having two coxswains, which is really no coxswain at all, because we don't know which one everyone else will listen to. It's not a beat; it's just noise.
"I’m not opposed to an NGDP target, I’m only trying to bring some realism to what we can expect from one, especially in a liquidity trap when fiscal policy is AWOL. No framework is perfect; NGDP simply has to be less imperfect than the alternatives."
Yep. That's all I would claim.