Scott Sumner writes

What interests me most is that [Pianalto] talks as if the Fed is still steering the nominal economy, despite near zero interest rates.  Other people may not see it that way, but swing voters at the FOMC certainly talk like they are still “doing monetary policy.”   In one sense that’s reassuring—we’d hate to see those at the controls claiming that the steering mechanism for the economy was stuck.  On the other hand it’s also a bit dismaying, as the marginal crew member of USS Nominal GDP seems happy with the ship’s course; even as Obama, Romney, Bernanke, 14 million unemployed, and the US stock market think it’s obvious that aggregate demand is too low.

I’d like to encourage a shifting of the conversation. I think virtually all nominalists agree that there is monetary policy at the zero lower bound.

What we disagree about is whether the Fed’s reaction function changes significantly in the presence of the zero lower bound. Indeed, that might be an interesting paper.

This could take on several forms. One, as I suggest is that the reaction function could become asymmetric, so that the Fed attaches a very high loss weight what we might call “non-Keynesian easing” but “attaches a low loss weight to “Keynesian easing.”

In this case normalizing interest rates though expansionary fiscal policy in effect serves to soften the stance of monetary policy.

There are at least two other hypothesis worth exploring:

One is that the Fed’s inflation weighting simply rises as it hits the zero lower bound. I don’t have an immediate explanation for this apart from the hypothesis above, but we should consider that something odd simply happens in the minds of central bankers as they bump up against the zero lower bound. The evidence suggests this.

The second is what I am loosely calling hysteresis in policy making. This is essentially the hypothesis that policy makers become inured to their past mistakes. So in a regime that has had persistently high inflation, the loss weighting of inflation begins to drop.

In a regime that has persistently high unemployment the loss weighting on unemployment begins to drop. Policy makers themselves begin to accept the new normal.

All of these cases, have important policy implications.

One of course is a strongly growing level target that assures that the Fed never bumps up against “non-Keynesian easing” the other is that fiscal policy or more importantly automatic stabilizers could have an important role to play early in a recession.

This is very out of the box, but imagine a standing Civilian Corp that works like this. At any time anyone can sign up for 6 months of work at the Civilian Corp. You will be paid 70% of your previous average salary and will not loose unemployment eligibility. Though you can only do one round in the Corp every five years or so.

The Corp then outsources you out to non-profits to do charitable work.

The point of the Corp is this: When a large unexpected aggregate demand shock hits  many people will choose the Corp over running out their unemployment benefits.

The huge rise in Corp employment will serve as an obvious sign to policy makers that they have failed. Yet, the sting of true jobless will not have hit yet. This gives a buffer zone where it is undeniable that monetary policy is falling but the worst of joblessness has not yet come about.

The hope would be that policy makers would move to shift monetary expectations quickly so as to prevent the emergence of true joblessness.

Now, I am fully aware that in the ideal monetary regime we never even get to this point because the Central Bank is always targeting levels. However, in practice the regime is likely to make some sort of unforeseen error and need to change expectations quickly.

This at least give us a buffer.

I am of course not hanging my hat on this idea, just throwing it out.

My real concern is measuring policy makers loss functions under different regimes.