というNBER論文をエガートソンらが上げている。原題は「The Slanted-L Phillips Curve」で、著者はPierpaolo Benigno(ベルン大)、Gauti B. Eggertsson(ブラウン大)。

A slanted-L curve is well-suited to represent the non-linearity of the celebrated Phillips curve. We show this using cross-country data of major industrialized economies since 2009, including the inflationary surge of the 2020s. At high unemployment rates, an increase in demand reduces unemployment without creating strong inflationary pressures. Meanwhile, supply shocks have a muted effect. At sufficiently low unemployment, there is a labor shortage, so that the economy is at full capacity. Then, higher demand is inflationary, and supply shocks are amplified. We derive a model of a slanted-L curve.

論文の著者はIt’s Baaack:2020年代のインフレ高騰と非線形のフィリップス曲線の復活 - himaginary’s diaryで紹介した論文と同じであり、そちらの論文の分析を各国に適用した形になっている。以下は1/17のエガートソンの連ツイによる解説(ungated版へのリンクも含まれている)。

From AEA P&P with @PierpaBenigno. Fit Phillips originally proposed curve (using prices not wages)
on cross country data from 2009 and (in black): A Slanted-L Phillips curve! Webcast from ASSA below 1/
This is just loose suggestive evidence. We add a model that gives the estimated slanted-L in blue in the figure. Link to paper: 2/
This short paper is build on our much more detailed statistical analysis and more detailed model focused on US:
One of our basic point in both papers is that the reason why most of us failed to forecast the inflation surge in 2020s is that we overlooked what was a conventional wisdom when Phillips wrote his paper. 4/
Once you hit production capacity inflation responds strongly to any excess demand. This conventional wisdom got lost in the fog of the Great Inflation of 1970's. There unancored inflation expectations played central role and the labor market was a side-show. 5/
So as inflation surged in 2020's people were greatly comforted by that inflation expectations, especially longer term ones, were very stable, forgetting the old conventional wisdom of the slanted-L that motivated Phillips. 6/
One underappreciated feature of the old conventional wisdom is that once you hit maximum capacity, i.e. hire everybody that wants to be hired, then it is not only demand shocks that have a bigger impact on inflation. 7/
Another important implication is that supply shock also start having a bigger effect. What we have been seeing recently is a fall in inflation largely due to supply constraints easing. 8/
But with the labor market still tight, it leaves the Fed vulnerable to shocks that will have a bigger inflationary effect than what we would expect based on past estimates. 9/
This makes the Fed's job pretty challenging going forward and highly data-dependent -- and probably more so that you would normally expect. end/

@PierpaBenignoと共著のAEA Papers and Proceedingsから。フィリップスが元々提唱した曲線

を2009年以降の各国データに当てはめ(黒線)(賃金ではなく物価を使用)、および、傾いたL字型のフィリップス曲線! ASSAでのWebcastは以下。