It’s not hard to figure out what went wrong with the Fed’s FAIT policy initiative of 2020. Right after adopting a policy containing the phrase “average inflation targeting”, the Fed allowed the average inflation rate to diverge sharply from the roughly 2% average we saw during the period of 1991-2020. They adopted an asymmetric policy of making up for inflation shortfalls, but not inflation overshoots. The result was predictable, and will likely cost Biden another term in the White House.
Former Chicago Fed president Charles Evans has a new paper that discusses what went wrong and how to fix it. In the end, he sees the Fed’s FAIT regime as being relatively sound, and views the recent monetary policy mistakes as being relatively modest. In my view, the policy tweaks he proposes would not be enough to fix the underlying problem. The Fed needs to take a more radical step, either make FAIT fully symmetrical, or (even better) switch to something like NGDP level targeting.
Evans does acknowledge that there was excess demand stimulus in 2021, but in my view he understates the just how far policy went off track. Much of the paper focuses on how supply problems contributed to high inflation in 2021 and 2022, a problem that turned out to be largely transitory.
If you take the longer view, it begins to look like the real problem was nominal—excessive growth in NGDP. Evans suggests that the trend rate of growth in real GDP is about 1.8%. That suggests that NGDP should grow at roughly 3.8% over the long run. Here are the actual growth rates over the past 4 years:
Real GDP, 2019:Q3 to 2023:Q3: 2.0%
Nominal GDP, 2019:Q3 to 2023:Q3: 6.2%
Thus essentially all of the excess inflation (for the GDP deflator) was due to excessive growth in demand, at least in an accounting sense. NGDP growth was 2.4%/year above the level required to hit the inflation target, a total overshoot of roughly 10% over 4 years. That NGDP overshoot is roughly comparable to the inflation overshoot over the past 4 years. Indeed, it’s surprising that inflation was not even worse, as the PCE index has averaged about 4% inflation over the past 4 years, a bit below the figure you might expect from NGDP overshoot. If in 2019 you had been told that we’d have 6.2% NGDP growth, no Covid, and no Ukraine War, what sort of PCE inflation rate would you have expected? Be specific.
In fairness, the excessive growth in demand may have caused RGDP to slightly overshoot trend growth. Slower NGDP growth would have shown up as both slower RGDP growth and slower inflation. But as long as NGDP growth stayed close to 4%/year trend line, I would have expected RGDP to return fairly close to its natural rate, once Covid was over.
More importantly, I think a lot of inflation targeting advocates overlook the fact that a level targeting approach makes a severe inflation overshoot much less likely to occur in the first place. There’s far too much discussion of whether the Fed should have begun raising rates at this meeting or that meeting, and not enough discussion of what sort of monetary regime would create stabilizing speculation in the financial markets.
With a 4% NGDPLT policy, market interest rates on Treasury securities would have risen sharply in late 2021, as participants saw that NGDP was likely to overshoot the trend line and require subsequent contractionary policy. That rise in rates would have made the NGDP overshoot much smaller in the first place. And as long as NGDP is on track, any inflation overshoots due to supply side factors really will be “transitory”.
Based on what I’ve been reading from various Fed officials, I believe the Fed will fail to incorporate this perspective in its upcoming 5-year review, and hence will fail to come up with the sort of policy regime that would prevent a repeat of the mistakes of 2008-09, or the mistakes of 2021-22.
HT: David Beckworth
Update: I recently did a podcast with Dan Schulz, discussing everything from art to economics. Here are some links:
- YouTube: https://www.youtube.com/watch?v=b9Ek9njaCIk
- Apple: https://apple.co/3GoEXlp
- Spotify: https://spoti.fi/47CB7kv
- Substack (with AI transcript): https://www.danschulz.co/p/scott-sumner
- Twitter (X): https://x.com/dnschlz/status/1730366633619730701