前回に続き、サマーズとクルーグマンツイッターでの議論を紹介する。今回は直接的な論戦ではなく、こちらのLaurence Ball(ジョンズ・ホプキンス大)、Daniel Leigh(IMF)、Prachi Mishra(同)*1による論文「Understanding U.S. Inflation During the COVID Era」を巡る両者の見解の紹介となる。


@jasonfurman points out careful econometrics of Ball & co-authors which corroborates the argument I have been making (contra @paulkrugman @federalreserve & team transitory) that bringing inflation down to 2% will likely require a significant recession.
Three things drive the results.
First, to gauge labor market tightness one needs to look at vacancies, not just unemployment. As @Domash @ojblanchard1 & I have stressed the outward shift in Beveridge curve implies an increased NAIRU and that the economy is now well overheated.
Second, careful analysis of inflation data focusing on median inflation suggests significant underlying inflation rate that will require slack to bring it down.
Third, there is only a limited sensitivity of inflation to slack. Even on an optimistic view about sacrifice ratios, a few point years of above neutral unemployment are necessary to bring inflation down by several points.
I am glad Chairman Powell and his colleagues are now being clear that they will do what is necessary to contain inflation. The more determined they appear, the lower will be the cost of reducing inflation.



So, I'm reading the Ball/Leigh/Mishra paper Jason Furman calls "the scariest economics paper of 2022." It definitely paints an ugly picture. And these are serious guys, not at all permahawks. Their grimness might well be right. But I have questions 1/ https://brookings.edu/wp-content/uploads/2022/09/Ball-et-al-Conference-Draft-BPEA-FA22.pdf
A lot of discussion out there about their use of vacancies to unemployment as a measure of labor market tightness. Indeed, the apparent outward shift in the Beveridge curve is the main driver of their pessimism. What do I think about that? No strong views 2/
Reasonable people on all sides here, and I don't feel that I understand post-Covid labor markets well enough to make a judgment. Put it this way: if we've had enough structural change to massively shift the B curve, we've also had enough to screw up standard measures. So, ??? 3/
Another issue is the measure of core inflation. They prefer median inflation, which I too was fond of for a while. But at this point, as I understand it, median inflation is basically housing. So is housing a good measure of underlying inflation? 4/
Or are we in part looking at a one-time adjustment to changes in work-life patterns? I suspect there's some of that. Obviously you don't want to throw out all the positive numbers. But wages, for example, suggest lower underlying inflation than their measure 5/
One interesting point in their paper is that they give only a small role to rising inflation expectations. This is very different from the 70s. They suggest that expectations may yet become unanchored in the face of high actual inflation, which is certainly a worry 6/
But right now we're in a period of *low* headline inflation, with falling commodity prices, easing supply chains, etc. Market-based expectations have been falling, not rising. 7/
Anyway, most of the outlook depends on how overheated we think labor markets really are; it's all about which Beveridge you choose to drink. And I honestly don't know what I believe 8/
欠員率と失業率の比を労働市場の逼迫度の指標として使う彼らのやり方については、多くの議論がなされている。実際、ベバリッジ曲線が外側にシフトしたとみられることが彼らの悲観論の主因となっている。私の見解はどうか? 特にこれといった見解は無い。
それとも、部分的には、ワークライフのパターンの変化における一度だけの調整を我々は目にしているのか? それもあるのではないか、と私は思う。すべてのプラスの数字を捨て去ることができないのは明らかだ。しかし例えば賃金は、彼らの指標よりも低い基調的なインフレを示している。


I agree with @paulkrugman that these are the right questions to ask. And I don't know the answers to them, but did my best in my comments (see next tweet).
And I'm wildly jealous of this line: "it's all about which Beveridge you choose to drink".
In my comments, described here, I assumed the Beveridge curve shifts 2/3rds of the way back to where it was prior to the pandemic. Why did I assume that? I have no conviction whatsoever, but I do think that shifting all the way back can't be a best guess.
P.S. One quibble with Paul: I think he was right in 2021 to advocate for the median CPI and I was wrong to be less excited about it. It is, indeed, often shelter--but not sure why we care because that only happens because half of prices increase more than that and half less.
ここに示す私のコメントでは、ベバリッジ曲線がパンデミック前にいたところまで2/3戻ると仮定した。なぜそう仮定したか? 確信は何も無いが、ただ、完全に戻るのが最善の推定とは思わない。


*1:cf. 同じ著者たち(+α)による関連論文