With the #JobsDay data there are criticisms about the increase in unemployment, but labor-oriented recession indicators, like the @claudia-sahm.bsky.social rule, Modified U6 #SahmRule, the #MichezRule, still do not point to a downturn.
A bad or weakening economy is not the same thing as recession.
Economists use the Sahm Rule to predict recessions: a smoothed unemployment rate increase of 0.5%. It has recently been triggered. #SahmRule #Recession #Economics
www.youtube.com/shorts/T8kXT...
When you look at the popular recession indicators: the #SahmRule, by @claudia-sahm.bsky.social, the #MichezRule, by Pascal Michaillat and Emmanuel Saez, which adds job vacancy data, and the U6 Modified Sahm Rule, by M. Sheppard, using underemployment data shows downturns sooner.
Using #SahmRule -style indicators with labor utilization shows a gradient: broader measures of underemployment (esp. U-6) flag rising slack—because vulnerable workers tend to be underemployed.
Web Interactive: markgsheppard.github.io/SahmRule/
🚨WP🚨 papers.ssrn.com/sol3/papers....
#JobsDay #EconSky
With respect to the "slow to hire/fire" rhetoric, data shows using unemployment by: race, u-measure and education, in a #SahmRule like function, by @claudia-sahm.bsky.social, shows some are consistently "first to be fired" in a downturn.
#EconSky #JOLTS
🚨New WP🚨 papers.ssrn.com/sol3/papers....
Using #unemployment rates disaggregated by labor utilization, in a modified #SahmRule, shows broader measures of underemployment indicate #recession months prior. In part because economically-vulnerable groups tend to be #underemployed.
🚨New Working Paper🚨 papers.ssrn.com/sol3/papers....
#EconSky
Using unemployment rates, disaggregated by education, in a modified #SahmRule, shows that low-educated groups experience recessions months prior.
Data shows unemployment rates for many economically vulnerable groups have been rising.
🚨New Working Paper🚨 papers.ssrn.com/sol3/papers....
#EconSky
Using unemployment rates, disaggregated by race, in a modified #SahmRule, shows that marginalized racial groups experience recessions months prior.
Data shows Black unemployment rates have been rising.
#EconSky
papers.ssrn.com/sol3/papers....
To put an even sharper edge to this: my research shows that when substituting in Black unemployment into a modified #SahmRule, the data has been cautioning of a downturn for some time.
Recessions are not uniform. The data shows downturns hit some people/areas/sectors first.
#EconSky
To put an even sharper edge to this: my research shows that when substituting in Black unemployment into a modified #SahmRule, the data has been cautioning of a downturn for some time.
Recessions are not uniform. The data shows downturns hit some people/areas/sectors first.
#EconSky
The #SahmRule and the Michez Rule (which also considers job vacancies) are not triggering. The U6 Modified Sahm Rule, which captures slack, is triggered. But more important to note is the direction these turning point functions are trending. #EconSky #NumbersDay
WP: papers.ssrn.com/sol3/papers....
My research basically takes the premise of the Sahm Rule and asks “there’s like a billion datasets for unemployment, is there an even better #SahmRule?” And I look across a range of parameters, and blah blah blah, and basically find that if you just use U6 it’s much faster/little noisier than U3.
Ben raises a good point about the gap, how you treat the @claudia-sahm.bsky.social rule.
Here’s my shameless plug for the conclusion of my WP: the #SahmRule shouldn’t be understood as a singular heuristic, but rather a really flexible labor-oriented turning point framework.
I created an interactive tool allows for different measures of unemployment in a turning point analysis framework, modifying the #SahmRule by @claudia-sahm.bsky.social, and the 20-24 group is seeing increases in their unemployment relative to the recent past.
markgsheppard.github.io/SahmRule/int...
I modified the #SahmRule, by @claudia-sahm.bsky.social, creating 3.2k spacial Sahm Rules, which show that #recession exposure can differ by place/time.
The bellwether of recessions are rich/white counties but they're slow; the fastest indicators are poor/black counties, but they're noisy.
When you compare the #SahmRule by @claudia-sahm.bsky.social, the #MichezRule by P. Michaillat and E. Saez, and the U6 Modified Sahm Rule, it shows that recessionary calls are premature but that the labor market has weakness particularly on the intensive margin. #JobsReport #EconSky
#Jobs? #Recession?
The #SahmRule signals a recession when the three-month average unemployment rate rises by 0.5 percentage points above its 12-month low. Right now, we’re at 0.17%, so technically no recession signal yet, but the trend is concerning.
With the #JobsReport out: no clear recession signal. The Sahm Rule, the Michez Rule, and my U6-Modified Sahm Rule (most sensitive to slack) haven’t triggered.
There is a impulse by some to look for recessionary conclusions, when the reality is the economy just sucks.
#SahmRule #MichezRule
When you look at the #SahmRule in a county-level spatial context, you realize that the midwest, from Minnesota to Ohio, are experiencing recession-like conditions.
Many policy economists are interested in the political implications of #tariffs, to me this is it.
#EconSky #JobsDay #JobsReport
The interactive #SahmRule tool I built based on @claudia-sahm.bsky.social's method—with flexible inputs & adjustable parameters, is great for identifying turning points.
At a glance, uncertainty and the yield spread are worsening, headline unemployment looks fine, but broader metrics raise concern.
The Modified Sahm Rule: Lunch Tray Edition
If food waste in Title 1 schools decreases by ≥20% from its 12-month average over a 3-month span,
a household-level economic contraction is likely already in progress. #SahmRule
Currently on a training mission with allies, should be back soon, very limited service until then. But I didn’t want to miss #JobsDay, so my code auto-fetches the data for the interactive #SahmRule tool.
Using the #SahmRule as a turning point framework for unemployment (U3), and mapping all 3,144 counties, shows that the counties that are historically the most accurate indicator of recession are the wealthy Virginia suburbs of D.C.
#EconSky
I've updated the beta version of the #SahmRule tool to include some non-employment measures, treating the rule as more of a parameterized turning point function, and it shows uncertainty/yield spread/consumer sentiment worsening, but the labor market is somewhat stable.
#EconSky
And just to add, if you put the U1-U6 into a turning point function, like a #SahmRule, it shows that unemployment is mostly stable.
So while real-time GDP forecasts are negative, most potential recessionary indication in the labor market seem premature.
#EconSky #JobsDay
The #SahmRule interactive tool, has been updated so you can add lines, and download data. The tool is still in beta, and there's some bugs to work, but the project is coming along.
markgsheppard.github.io/SahmRule/int...
Among other things, I'm working on a Beta V.2 of the #SahmRule interactive tool that will allow for adding lines, for comparison, and allow users to download data.
Generally, I think researchers could be better about the degree to which their results are downstream of parameters.
With the interactive tool on the #SahmRule, maxing-out the smoothing parameter tells a pretty clear story; the noise zeros-out, but the directionality remains, which is what matters.
Most attention is paid to U3 of the #JobsReport in trying to assess the macroeconomy, but the U6 does a much better job of capturing marginal changes in slack.
The WP I've been working on that modifies the #SahmRule, shows the U6 Sahm Rule is pushing up.
#JobsDay
@claudia-sahm.bsky.social
🚨 NEW INTERACTIVE RECESSION TOOL 🚨
The #SahmRule is the best tools we have to indicate a recession. I extend the logic laid out by
@claudia-sahm.bsky.social, building off work by
@pmichaillat.bsky.social and Saez, to create an interactive tool. #EconSky
markgsheppard.github.io/SahmRule/int...
