I am writing this post following our presentation at the Jackson Hole Symposium. You can find the latest version of our paper at this link, where we revisit the recent inflationary surge and its subsequent decline through the lenses of the Phillips and Beveridge curves. This period has been extraordinary in many respects, as we discuss extensively in the paper.
With inflation being tamed, the market has shifted its focus to concerns about an impending recession. The well-known Sahm rule, along with recent enhancements by Michaillat and Saez (2024), is signaling the possibility of a recession, with the latter even suggesting it could have begun as early as March 2024. Given the Sahm rule's strong track record, these signals are indeed cause for concern.
However, I argue that it is premature to worry about an imminent recession. Moreover, even if these concerns were justified, there are actions the Fed can take —— specifically, aiming to keep unemployment close to the BT unemployment rate.
The figure above illustrates the unemployment rate since 1947, with NBER-defined recessions shaded in grey. It is evident that recessions are typically accompanied by rapid increases in the unemployment rate, rather than necessarily a high unemployment rate itself. This observation underscores why the Sahm rule, which compares the average of the latest three unemployment figures with the minimum unemployment rate over the previous 12 months, is effective in capturing this rapid surge —— this is why it works most of the time.
The same figure also highlights periods of particularly tight labor markets, shown in the rosy areas, defined by a vacancy-to-unemployed ratio exceeding the Beveridge Threshold (BT). For the U.S. economy, this threshold is well identified by a unitary value—when the number of job seekers equals the number of vacancies. Such tight labor market periods are rare in U.S. history, with only four occurrences since 1947: during the Korean War in the 1950s, the Vietnam War in the 1960s, and more recently, the pre-COVID and post-COVID periods. Historically, similar tight labor markets have also been associated with wartime, such as during WWI and WWII.
Notably, all U.S. inflationary episodes, except for the 1970s, have been associated with periods of tight labor markets. However, the concerning insight from the figure is that all recent periods of labor market tightness have culminated in a recession. While this visual link suggests a possible causal relationship, it should be approached with caution. For example, the labor market tightness in the pre-COVID period had no causal link to the COVID-19 pandemic, which was the actual trigger for the ensuing recession.
The Beveridge-threshold (BT) unemployment rate provides insight into the level of unemployment below which the labor market becomes tight, historically leading to significant inflationary pressures. This metric is derived using JOLTS data, so it is only available starting from December 2000. The figure above presents the BT unemployment rate compared to the actual unemployment rate in the top panel, while the bottom panel illustrates the gap between the two.
Interestingly, the BT unemployment rate remained consistently below 4% before the Great Financial Crisis, even dropping to 3% at one point. However, during the financial crisis, it rose above 4%, never exceeding 4.5%, and eventually returned to just below 4% before the COVID-19 pandemic. During the pandemic, this rate spiked to 8%, gradually decreasing to 4.42% by its most recent observation in June 2024.
There have been only two instances in this period where the unemployment rate fell below the Beveridge threshold. These instances coincided with times when the job vacancy-to-unemployed ratio exceeded one. In the first instance, from 2018 until the pandemic, the gap between the unemployment rate and the BT unemployment rate did not drop below -0.5%. However, the post-pandemic period that began in May 2021 saw the gap suddenly widen, reaching a maximum of -1.08% in March 2022, which coincided with the peak of the inflationary surge. At that time, the unemployment rate was 3.65%, while the Beveridge-threshold rate lagged behind at 4.73%. This discrepancy is crucial for understanding the recent inflationary surge in the U.S.
This analysis reveals several important insights. The gap between the actual unemployment rate and the Beveridge-threshold (BT) unemployment rate is closing, indicating that the labor market is approaching a state of neutrality where inflationary pressures are subdued. Notably, this gap is closing at a different BT unemployment rate compared to the pre-COVID period: 4.42% now, compared to 3.89% previously. This change is largely due to a decrease in matching efficiency within the labor market, driven by the reallocation of demand across sectors during and after COVID.
It is, therefore, not surprising that the Sahm rule is flashing red, given that unemployment dropped to a historically low 3.4% in April 2023 before rising to the more neutral level of 4.3% in July 2024. While the 3.4% unemployment rate indicated a significantly overheated market compared to the BT unemployment rate, the current 4.3% suggests a return to more balanced, neutral conditions. Taken together, these developments point toward a return to normalcy rather than the onset of a recession.
Another important observation is that all adjustments should be influenced by policy, particularly monetary policy. As labor market conditions normalize, the Fed must carefully assess whether its contractionary measures on the unemployment gap have not been excessive. The goal should be to guide the unemployment rate to a level that is above but close to the BT threshold, thus avoiding the risk of renewed inflationary pressures from demand and supply shocks and steering clear of the rapid increases in unemployment that have characterized previous recessions.
Policymakers should take heed. This time could indeed be different, with the Sahm rule potentially signaling a false positive.
References
Pierpaolo Benigno and Gauti Eggertsson. 2024. Revisiting the Phillips and Beveridge Curves: Insights from the 2020s Inflation Surge. Jackson Hole Economic Policy Symposium, August 2024.
Pascal Michaillat, Emmanuel Saez. 2024. Has the Recession Started?