I am getting increasingly annoyed by all of the discussion of whether the Fed will succeed in getting inflation back to 2%. That should not be the Fed’s goal.
In 2020, the Fed instituted a policy of “flexible average inflation targeting”. Unfortunately, the exact nature of that policy is somewhat ambiguous, but under no reasonable interpretation could “getting back to 2% inflation” be viewed as consistent with FAIT. Inflation is currently running far above 2%, and is expected to continue to exceed that rate for at least the first half of 2022. If FAIT is to mean anything, the Fed needs to shoot for an inflation rate well under 2% during the mid-2020s.
Now you might argue that pushing inflation down to say 1.6% is not consistent with full employment. I doubt that. In the late 2010s, we had the lowest unemployment in 50 years and PCE inflation ran slightly 2%.
But given the fact that the Fed adopted a 2% average inflation target, they should try to hit that target. If Jay Powell doesn’t believe that low inflation in the mid-2020s is consistent with full employment, then the Fed never should have adopted FAIT in the first place. Instead, they should have adopted something like NGDP level targeting. Unfortunately, the Fed went in a different direction, and so now it needs to establish credibility by hitting its target.
The term “transitory” can have one of several meanings. Under ordinary inflation targeting, transitory inflation means an inflation rate temporarily above 2%, before reverting back to 2%. But under average inflation targeting, transitory inflation requires a period of above 2% inflation be followed by a period of below 2% inflation.
The TIPS markets seem to be forecasting above 2% inflation over the next 10 years. That’s bad news for Powell. Monetary policy is currently too expansionary.
READER COMMENTS
Jose Pablo
Jan 7 2022 at 1:07pm
Why is so magical about the 2%? (whether as an “ordinary inflation target” or as a an “average inflation target”).
Why not 0% or 4%?
Scott Sumner
Jan 7 2022 at 2:06pm
There’s nothing magical about 2%. But that’s the figure the Fed chose.
What’s so magical about a 65mph speed limit? Why not 63mph? Why not 67mph?
Still, when I drive I look at the actual speed limit when making decisions.
Christophe Biocca
Jan 7 2022 at 2:35pm
Nothing beyond it being the declared target, I think. A Fed that reacts to overshooting its 2% target by saying “actually we’re aiming for 4%” is a Fed you can’t trust to stick to 4% either.
Scott Sumner
Jan 7 2022 at 3:57pm
Exactly.
Lizard Man
Jan 9 2022 at 6:47am
I disagree. If the Fed wanted to raise its target for inflation, how can they do so while minimizing political backlash? They themselves might not even know that they want to raise the target until it has actually happened. My own belief is that a higher inflation target than 2% is better at ensuring full employment, and that because of scarring from unemployment, failure to maintain full employment has a much higher negative impact on long term economic growth than higher inflation. I don’t know, but do wonder if the Fed looks at disability claims and labor force participation after the Great Recession and agrees.
I think that the current recovery provides some evidence that higher inflation (and hence lower real wages) boost employment. But it is also evident from the TIPS data that choosing this tradeoff isn’t consistent with an AIT of 2%. But could the Fed have raised their target inflation rate to say, 3.5% without politicians going ape on them in a way that the Fed views as imprudent? And any sudden change from a 2% target would include a big loss of face for longer serving Fed members. If Powell does want to raise them target rate, higher inflation now while guiding it down to a higher baseline level while new board members turnover would be a sensible strategy. Then the new members can be responsible for lowering inflation to 3%, not raising it to 3%.
Scott Sumner
Jan 9 2022 at 1:17pm
I don’t know of any evidence that raising inflation above target leads to anything more than a short-term boost in employment, at the cost of lower employment in the future.
Mark Brophy
Jan 7 2022 at 4:29pm
The Fed should not try to establish credibility by hitting its target because they don’t have the power. Inflation will be brought under control when that’s the desire of the voters. They will inflate away the debt for about 10 years and create a stable currency later.
Jon Murphy
Jan 7 2022 at 5:57pm
It’s been a while since I last voted, but I don’t recall a referendum on inflation on the ballot
Mark Bahner
Jan 8 2022 at 11:34am
Is it even possible to inflate away the debt with so many federal benefits tied to the inflation level?
Also won’t interest rates on the federal debt go up when it becomes clear the federal government is trying to inflate away the debt?
stoneybatter
Jan 7 2022 at 7:07pm
Scott, when thinking about the 10y TIPS breakeven (currently at 2.46%), I have two points:
1) CPI runs ~30bps above PCE, so it’s actually not that far above target
2) the exact level of the breakeven (ignoring other considerations like liquidity effects) is the mean probabilistic expectation. That is, there is likely something like a 75% chance of exactly 2.3% CPI (2% PCE) inflation, a 5% chance of an undershoot at 1.5% CPI inflation, and a 20% chance of an overshoot at 3.3% CPI inflation. That results in 2.46% mean expectation. This is consistent with pricing from other markets, like inflation caps and floors, which show a relatively high risk of an overshoot and a relatively low risk of an undershoot.
I would argue that the Fed should still be tightening policy this year, ideally until the market is pricing symmetrical tails, but it’s different than saying the market is pricing outright failure.
Scott Sumner
Jan 8 2022 at 12:14pm
“Failure” is too strong a term. Obviously inflation of slightly above 2% is better than 5% or 10%, but inflation of 1.8% would be better yet.
Thomas Lee Hutcheson
Jan 9 2022 at 6:11am
Doesn’t the optimal target for inflation (or NGDP) depend on the “amount” of stickiness of prices and the expected “size” of the relative price changes that inflation permits? How should the Fed figure out whether 1.8% or 2.0% or something greater. [I am not talking about arguments higher nominal inflation so as to avoid the Zero Lower Bound.]
Thomas Lee Hutcheson
Jan 7 2022 at 9:59pm
I think the Fed needs to do something, maybe even get actual inflating below 2% PCE for a few quarters, to get expectations back around 2% PCE over time. If they think prices are so sticky that an average of 2% PCE is not compatible with full employment (but how will they know until they try?) they should move the target up a bit.
MarkLouis
Jan 9 2022 at 8:25am
Even if wages are sticky, why would we see layoffs at a time of record corporate profits in response to slightly below 2% inflation? Something doesn’t add up – are we forced to forever target high and rising profit margins?
Michael D Sandifer
Jan 7 2022 at 10:25pm
I don’t think the Fed needs to honor it’s explicit target to be credible. It just needs to be consistent. Also, the dual mandate gives them more flexibility.
There’s no reason the Fed couldn’t call this FAIT, but level target NGDP instead. If they do it consistently, they have the kind of credibility that matters most.
Jon Murphy
Jan 8 2022 at 8:35am
I’m not following you: what good is consistency without an announced target? If they announce they are targeting X, but in reality they are targeting Y, then the fact they cannot hit X looks like 1) inconsistency and 2) lack of commitment. Why not just announce you are targeting Y?
Michael D Sandifer
Jan 8 2022 at 11:13am
Jon Murphy,
I would turn that around and say, “What good is an announced target if they don’t hit it with any consistency?”. Which is more important, the rhetoric, or the behavior and results?
Jon Murphy
Jan 8 2022 at 11:23am
That’s not really a turn around of my question. You posited there is some goal they can consistently hit. The variable is not consistency, but rather credible commitment. If there is a target one can consistently hit, what good is it if you don’t announce the goal? Consistency is useless if there’s no way to measure
Michael D Sandifer
Jan 8 2022 at 12:11pm
Markets can predict the Fed’s behavior better than the Fed itself, as they’ve shown time and again.
Jon Murphy
Jan 8 2022 at 12:40pm
Ok, but that’s because the Fed announces their goals.
Again, I’m not seeing your point
Michael D Sandifer
Jan 8 2022 at 2:05pm
Jon Murphy,
My point is that for many years after the Great Recession, the Fed claimed to have a 2% inflation target, while consistently undershooting that target and consistently issuing explicit inflation forecasts that overestimated future inflation rates. Markets though reflected below-target inflation expectations over that period.
The explicit messages the Fed tries to send are not what markets care about.
Jon Murphy
Jan 8 2022 at 2:36pm
These two sentences are contradictory. Markets did adjust to the (moderately) below 2% inflation. But they were able to adjust because the Fed’s goals were explicit and consistent. Markets could make reasonable predictions about Fed policy.
So, again, I go back to my original question: consistency only matters insofar as expectations are properly set. If the Fed states that they’re targeting a 2% inflation rate, but then targets NGDP, they will look inconsistent and unreliable. Markets would have a harder time adjusting. Why, then, not simply state that they’re targeting NGDP?
Consistency is only a virtue if one is not a screw-up. Not properly communicating one’s goals makes one look like a screw-up.
Scott Sumner
Jan 8 2022 at 12:16pm
I agree that a NGDPLT targeting approach would be consistent with FAIT, but policy is also currently too easy using that metric.
Michael D Sandifer
Jan 8 2022 at 2:07pm
Scott,
Yes, money’s too easy under a 4% NGDPLT, such as you favor, but not under, say, a 5% target.
MarkLouis
Jan 9 2022 at 8:29am
The Fed went to great lengths to design and roll out FAIT. If a few years later the Fed scraps it, this basically tells me they have no idea what they are doing.
Brian
Jan 8 2022 at 7:37pm
I think it is reasonable to say the average is to be calculated for the period that starts on the date the averaging policy is announced and ends on the date that it is being calculated with the proviso that a reasonably long period of time has elapsed. Regardless of what I think is reasonable maybe the average is to be calculated for the period that started near the GFC, such as December 2007 when an earlier recession started or March 2009 when the stock market bottomed?
Kevin Erdmann
Jan 8 2022 at 8:36pm
If they do bring it back down to 1.6%, then will they have to commit to hitting 2.4% after that?
Scott Sumner
Jan 8 2022 at 10:14pm
No, once the average rate of inflation returns to 2%, they target 2% from that point forward.
Grant Gould
Jan 9 2022 at 10:13am
If the “average” in FAIT were a simple uniformly weighted sliding window average, they absolutely would — this sort of “washboarding” oscillation is the reason that at least in my lines of work (telecoms, robotics) nobody uses sliding window averages like that.
I think it is reasonable to assume that the Fed has at least one competent control theory mathematician on board somewhere and is internally using exponential weighting or some similarly damped average.
Michael Rulle
Jan 12 2022 at 8:34am
My problem on time frames continues to grow.
While FAIT seems right——-it appears to be undefinable. What is the time frame? At some point the Fed’s target was 2%. But on average inflation was slightly lower——-which seems correlated in time (2019?) with the decision to go to FAIT. Unless 2019 was the starting point—-which Scott has said before. And then comes Covid. Now what? How fast should we get to a 2% average? Is that defined?
Failing on a target is presumably bad because actions are taken to hit it which turned out to have been the wrong actions——and therefore must have a negative impact on the economy —-or one would not care. Hence one tries to make up for it ——in some way that is not intuitive. It feels like the sunk cost fallacy.
If we were to go to targeting NGDP, wouldn’t we have the same issue? Perhaps it is the the attempt at precision is pursuing yet another fallacy ——like many other models in economics. It seems possible to have something like FAIT or NGDP targeting to be a guide post——without presuming failure if missed by .2%.
Comments are closed.