The recent JOLTS report shows a drop in the number of vacancies and an uptick of the quits rate. This note updates the results of our models that use the ratio of vacancies per unemployed (V/U) and the quits rate (Qr) as measures of slack. The conclusion remains the same of last year: V/U and Qr need to fall to around 0.4 and 1.7, respectively to hit target in 2025. Translated: a non-mild recession.
Our 2022 notes
In 2022 the model(s) correctly predicted the evolution of core PCE price inflation. In March and August 2022 we released two notes (here and here) in which we estimated that disinflating the US economy required a significant fall in V/U or Qr, possibly much more than argued by Waller and Figura (2022). Not only, but our estimates showed that: (i) core PCE price inflation was expected to remain at or above 3 percent in 2023, and (ii) it was very hard to see core inflation at target in 2024, unless a severe recession hit the US economy. Considering that the first note was written in March 2022 and the models delivered an accurate forecast (against consensus and professional forecasters), we take the models seriously.
What the model(s) say now
It takes a (non-mild) recession to hit target in 2025. We use our “main” model modified to include V/U or Qr as a measure of slack. We run a baseline and two alternative scenarios. In all scenarios V/U and Qr fall from the current levels, reflecting the expected slowdown of the economy. In the baseline, V/U and Qr reach 1 and 2.2 respectively at the beginning of next year and stay there in 2025. These levels are the averages of the variables at the end of 2017 when slack was estimated to be roughly zero. On the other hand, in scenario1 we assume a mild recession which brings V/U and Qr to 0.8 and 2.0, roughly consistent with the levels of the variables in 2015. Finally, in scenario2 we assume a deep recession which brings V/U and Qr to 0.4 and 1.7, their 2013 averages respectively.
Table 1 and Figure 1 show the evolution of core PCE price inflation (YoY) in the baseline and in the two alternative scenarios. In all scenarios core inflation remains elevated this year at around 4 percent. Going beyond this year, the models project core inflation between 3 and 3½ percent at the end of the medium-term in the baseline. Under scenario1 core inflation is expected in between 2½ and 3 percent in 2025 depending on the measure of slack used. Reaching target at the end of the forecast horizon, as mentioned, is achieved only in scenario 2.
Table 1. Model-based forecast of core PCE price inflation (YoY)
Note: the table shows the evolution of core PCE price inflation forecast (YoY) in the baseline and two alternative scenarios. We use two measures of slack: “Qr” stands for “Quits rate” and “V/U” stands for “Vacancy per Unemployed”. In all scenarios V/U and Qr fall in the forecast horizon but: (i) in the baseline they reach 1 and 2.2 respectively at the beginning of 2024 and stay there in 2025, (ii) in scenario 1 they reach 0.8 and 2.0, respectively, and (iii) in scenario2 they reach 0.4 and 1.7 respectively.
Figure 1. Model-based forecast of core PCE price inflation (YoY)
Baseline – using V/U
Scenario1 – using V/U
Scenario2 – using V/U
Baseline – using Qr
Scenario1 – using Qr
Scenario2 – using Qr
Conclusion and implications for the Fed staff
The downward tick in V/U in the latest JOLTS report has been interpreted by some as a sign that disinflation is here. Our research suggests that while it is a step in the right direction, the road is still long because the starting point is so high that the labor market remains very tight. In our experience, the favorite measures of slack of the Fed staff is the unemployment gap, not Qr nor V/U. Having said that, no matter what measure of slack we consider, we always reach the same conclusion: a significant slowdown of the economy (in fact, a recession) is needed to tame core inflation.