"Are we at full employment yet?"

I vaguely remember having seen this movie before, the earlier British version. And it's a frustrating movie to watch, because nobody knows where full employment is, except maybe when we've driven past it and can see it in the rear-view mirror. And the question itself always gets confused between asking where full employment ought to be, according to someone's map of an ideal world, and where it actually is, given the world as it is, and what can be achieved with aggregate demand policy alone.

If deviations of inflation from target responded instantly and consistently to deviations of employment from full employment, we wouldn't need to ask the question. Just keep inflation on target, and employment will look after itself. And if you think that "full employment" thus defined isn't ideal enough for your world, then OK, tell us what microeconomic (non aggregate demand) policies you would like to use to try to improve things.

But deviations of inflation from target don't respond instantly and consistently to deviations of employment from full employment. Inflation is a lagged and noisy signal. And inflation targeting itself, it seems to me, has made inflation an even more lagged signal, with an even lower signal/noise ratio. Divine coincidence doesn't work very well, and basing your policy on it seems to make it work even less well. So we still ask "Are we at full employment yet?" in order to try to keep inflation on target.

The main cost of overshooting full employment is not the inflation that results from doing so. The main cost is the cost of the extra unemployment when you bring inflation back down again. I've seen that movie before, in the 1980's, and I don't want to see it again.

Set aside the other benefits of switching from an inflation target to an NGDP level path target. If switching targets meant we wouldn't need to ask "Are we at full employment yet?" in order to figure out whether monetary policy is too tight or too loose, that would be a major advantage. Because it's a question we can't answer until it's too late. Instead we would simply hope that we are not at full employment yet: we would hope that target NGDP growth would in future be composed more of real GDP growth and less of inflation. And we would be forced to concentrate instead on microeconomic policies that might improve that composition.

Comments

[Scott Sumner]

Excellent post. I completely agree.

[John Chant]

Nice post with good message. Unfortunately, memories of the 80s are dimming.

[Nick Rowe]

Thanks Scott and John!

[John Chant]

Nick:

A second thought. If the main cost of inflation is the cost of bring it down, why bother?

. I think inflation has substantial costs in addition to the cost of bringing it down.

[Nick Rowe]

John: if the cost of bringing inflation down were the *only* cost, then I agree: why bother? But if it's just the *main* cost, and there are costs to inflation per se (as I agree there are), we would get a one-way ratchet of ever increasing inflation (which would be bad) if we never bring it down after a missed target.