Scott Sumner  

The Fed should now be aiming for below 2% inflation

Ralph Raico, RIP... Ralph Raico, RIP...

The Fed has a mandate to stabilize prices and provide for high employment. The Fed interprets that mandate as calling for 2% PCE inflation and an unemployment rate close to the natural rate. Tim Duy points out that the unemployment rate has recently fallen below the Fed's estimate of the natural rate:

On one hand, the unemployment rate plunged 0.3 percentage points in November to 4.6 percent: . . .

This is below the range of the longer-run central tendency (4.7 - 5.0 percent), sufficient to prompt a preemptive rate hike in December without dissent.

For most of the past 8 years, the unemployment rate has been above the natural rate. How should policy change as inflation falls below this benchmark?

Here's it's easiest to start with the case where we have a simple inflation target---2% inflation, regardless of what's going on in the economy. Under that regime the policy choice is simple---always aim for 2% inflation. Now let's add a second goal, unemployment close to the natural rate. In that case, you still aim for 2% inflation when unemployment is at the natural rate. But when unemployment is high you adopt a bit more expansionary monetary policy than would be called for by a strict 2% inflation target, and vice versa. This means the Fed should aim for slightly above 2% inflation during high unemployment years, such as 2008-15, and slightly below 2% inflation when unemployment is below the natural rate---like right now, or during the housing bubble.

Is this the Fed's actual policy? No. As far as I can tell, the Fed aims at something closer to the exact opposite---below 2% inflation during periods of high unemployment, and above 2% inflation during booms. On the face of it, this seems to violate the dual mandate. But since neither Congress nor the economics profession seems to notice this problem, there is no pressure to fix the policy. One reason I favor NGDPLT is that it would force the Fed to move toward a policy that it theoretically has already adopted, but in practice does not adhere to. A policy that is consistent with the Fed's legally mandated policy goals.

PS. Don't assume the takeaway from this post is about what the Fed should do today. The Fed's problems run far deeper than today's decision, which is relatively unimportant. Rather the Fed does not have the correct regime in place to deal with a turn in the business cycle. As long as we continue in this expansion, monetary policy will probably be roughly OK, and not a major concern for the economy. Our problems lie elsewhere. The problem is procyclical inflation.

Comments and Sharing

COMMENTS (10 to date)
Tom M writes:

The Fed should not focus on unemployment and the natural rate, two measures which are basically estimates for labor market health. If the Fed's inflation target is 2% and they are still below that target, they should not care if the unemployment rate falls below the "natural rate".

Scott Sumner writes:

Tom, I'm describing actual Fed policy targets. Obviously I favor a NGDP target. But if they are going to have a dual target of inflation and employment, then they need to try to actually hit their target. Do do otherwise would make things even worse.

Michael Byrnes writes:

Good post.

This is off topic, but what do you make of the unemployment rate being below the Fed's estimate of the natural rate?

EB writes:

The problem is procyclical inflation. Wow. If at least someone could define what it means.

Despite large deficits for a long time there are not inflationary pressures because of the low marginal cost of debt financing. The fiscal problem is huge waste both in collecting tax revenue and in spending tax revenue and borrowed funds. If the Fed were closed (that is, if its operation of a sort of a marketing board to control interest rates were terminated), only some economists would miss it.

EB writes:

Right now, WSJ's big headlines are

Investors Await Fed Decision on Interest Rates
Markets seek clues on next steps; Fed also set to release economic projections

After decades of pressing for a role in public policy, economists have succeeded in getting big headlines about changes in something called "monetary policy". But the emperor has no clothes.

Scott Sumner writes:

Michael, That's not surprising, as the labor market has been healing for many years now. It's only slightly below their estimate.

Brian Donohue writes:

The premise that we are at or beyond full employment is flawed.

We know this because wages still aren't going up a lot and job growth continues at a rate higher than the rate of new workers entering the labor force (2.25 million new net jobs over the past 12 months.)

The only explanation that makes sense is that discouraged workers continue to re-enter the labor market.

We never even hit the 2% target over the past several years, despite the logic of unemployment suggesting we should have been shooting higher.

But now we're all in a hurry to batten down inflation? I don't get it. There is room for jobs to grow yet, prolly 1-2 million discouraged workers to lure back in.

Aaron writes:

fwiw a Canadian politician (who might become opposition leader) flirted with a 0% target.

David N writes:

"...during the housing bubble." Are you kidding? Haven't you spent the last decade telling us EMH is everything and there's no such thing as a housing bubble?

EB writes:

Last night Greg Ip wrote this column

to explain why Trump may consider reappointing Yellen at the end of her term: their stars are aligned. He ignores the possibility that once in office Trump finally understands how irrelevant the Fed is despite all the talk about monetary policy. There is nothing that the Fed can do either to facilitate or to obstruct Trump's policies to make America great again (yes, I don't know his policies but neither him nor Yellen, Stan & Co. are idiots). But we can bet that if Trump's policies failed, the Fed would be the convenient scapegoat regardless of what the Fed's had been doing.

Comments for this entry have been closed
Return to top