A commenter on yesterday’s monetary policy post explains how NGDP targeting works:

The process is iterative.

In t=1, the Fed says “we want to get back to the level in one year. Therefore we will start by buying $100b of whatever this month, and check in in a month and see if that gets us there.”

In t=1+1month, the Fed says “OK, checking in. Looks like $100b/mo doesn’t quite get us there. $150b this month it is!”

The key is – as long as you think the Fed really will buy as much whatever as it takes to hit the target by the date, then you will act accordingly.

The thing about nominal growth is that it could be, in theory, all inflation. So RGDP growth over a certain timeframe could, theoretically, be zero or even negative while NGDP growth is hitting a target simply because the Fed is printing so much money.

So if the Fed can hit an NGDP target through purely inflationary money printing regardless of RGDP, it is credible that they can hit whatever NGDP target they want. And because they’re credible, you will act as though NGDP will be at the level in future periods and act accordingly, and that “acting accordingly” will tend to increase RGDP growth, meaning that you don’t actually HAVE to inflate very much to hit the NGDPLT.

I understand this argument. If the central bank can credibly claim that they will continue to buy bonds in ever-increasing numbers until the NGDP target is hit, and that nothing whatsoever will dissuade them, then yes, it makes sense that the market would expect that it will actually do that, and act accordingly.

But how do you earn that credibility? Presumably, the central bank would have to do something to demonstrate that steely-eyed resolve. What would it take?

Presumably, again, what it would take would be a willingness to persevere in its course in the face of substantial evidence that its policies were having deleterious effects. If the policy worked easily, after all, the market might conclude that the policy would only be followed if it worked easily.

So: the central bank announces an open-ended QE, much larger than previous rounds. It announces that it will not curtail the program if inflation goes above 2%, nor if real growth stalls – it will continue to buy bonds, in ever-increasing numbers, until the price level is restored to its previous path, and it is targeting to achieve this in a year.

Now: what happens if, after one quarter, inflation ticks up – but real growth drops, unemployment goes up, and business investment is in free-fall? Alternatively, what if inflation hasn’t ticked up much, growth remains anemic, the yield curve remains flat, and the financial markets look very frothy, leading to fears of a new financial crisis? (Which is where the Bill Grosses of the world think we are today.)

In either case, it would look like the policy was failing. And in either case, to remain credible, the central bank would have to continue in its course. It would have to say: no, our goal is reflation above all, and we are not going to be dissuaded by any short-term hiccups. In the long run, this is right policy, and we’re going to follow it no matter how bad things get.

After a couple of years of that kind of steely determination, yes, the financial markets would probably conclude: the Fed is serious about this. They are not going to permit deflation. They will always reflate, aggressively. So we don’t ever have to worry about deflation. At which point, deflation will become much less likely.

That’s pretty analogous to the situation Paul Volcker found himself in when he was appointed to chair the Fed. He settled on a policy of aggressively fighting inflation, and basically ignoring the other economic consequences of his policy. Unemployment went through the roof; America entered into a deep recession. Many (rightly or wrongly) trace the deindustrialization of America to this period; certainly, Volcker’s policy had a significant impact on the currency, and hence on patterns of trade. In 1982, this policy looked like a failure, and the Reagan Administration begged for relief. They didn’t get it – and suffered through a massive mid-term repudiation at the polls.

And then it started to work.

In the real world, the political circumstances that would permit an analogous reflationary policy to be adopted are fairly extreme. They certainly don’t persist today. There is substantial opposition, elite and popular, to QE even though there has been no uptick in inflation.

Moreover, even if the markets become convinced that monetary policy will produce Goldilocks conditions, that has a downside as well. Remember the “Greenspan put?” This was the conviction, based on Greenspan’s record at the Fed as a whole but strongly underscored by his brokering of a solution to the Long Term Capital Management situation, that a major financial crisis would never be permitted. Anything that threatened the fundamental stability of the system would lead to rapid Fed easing. So nobody need worry seriously about systemic risk – the Fed had that covered.

We all remember how that worked out, right?

In any event, my points about the inefficiencies of QE relative to the normal open-market operations of the Fed just mean that we should expect QE to have more negative side-effects than Fed easing normally does. In which case, it will be that much more expensive, politically and economically, to demonstrate the kind of resolve necessary to establish credibility.

All of which does not mean that we should raise rates. From my perspective, it definitely means that we should look forward to a cashless society where negative nominal rates are possible. And I would also argue it means we need to think about how the fiscal side can help the monetary side out, since both sides have associated costs and inefficiencies.