Thursday, 12 August 2010

Learning and expectations of NGDP

A response to Niklas Blanchard's post supporting Scott Sumner's NGDP level targeting proposal (also posted as a comment on that page).

My thoughts on this subject are not complete, but I've been thinking recently about how expectations are formed. I wonder about the following scenario:

Let's say the Fed does adopt an NGDP level target with a 5% annual increase, and this becomes accepted as the new orthodoxy. How will future wage negotiations proceed?

As an employee, I'd say to management: presumably you are good managers, and you expect to achieve growth in the company's revenues, profits and productivity at least in line with the economy as a whole. (I've yet to meet a manager who'd admit that they are not as good as the national average manager).

Therefore, you'd expect your top line revenues to increase by at least 5%, given a fixed amount of labour in your company. What's more, now that the Fed has successfully achieved a 5% NGDP increase for the last three years, I can confidently expect my cost of living, based on my reasonable expectation of a rising living standard, to increase by about 5% each year.

I would therefore suggest that my salary should incorporate a fixed 5% escalation every year from now on. What do you say, dear boss?

This may or may not be the scenario that plays out in reality, and the negotiating skills of management and employees will vary. But it's at least plausible that a 5% annual increase will become the new standard of nominal wage rigidity. And the consequence may well be a default 5% annual price increase for the majority of products and services. Competitive pressures will operate of course, but against a background of 5% increases across the board. If so, then we are back in a situation of sticky nominal wages and prices - it's simply a sticky 5% increase instead of a sticky 0% increase.

If this happens, will the macroeconomic benefits of the steady upward NGDP path disappear? Will we end up back in the same situation as the last few decades - where the policy target broadly achieves balanced growth for eight years and then we get a two year recession when investors become overconfident and it fails?

Maybe not - perhaps there is something psychologically special about 0% increases - but I am not so sure.

I'm still a supporter of the Sumner proposal, but I raise this as a cautionary question. In short, is any policy target bound to lose its effectiveness once people learn to expect it?

No comments: