Can the percentage of 25-54 year-olds employed tell us anything about recessions? (Part 2) And is the current decline going to be the first without a recession?

Click here to read Part 1 from Aug. 2024. In that post I looked at what happened to this percentage in the months leading up to a recession and also looked to see when a recession happened after a certain number of months in a row of this percentage rising.

In 2024 the 12 month average was 80.717%. In 2025 the average was 80.600% (just an 11 month average due to no data for October instead of 12). Even if we gave October 80.7% (which was the highest for any other month of 2025), the average would be 80.608%. That means that the the percentage of 25-54 year-olds employed was lower in 2025 than 2024.

 

In this post I look at all the 12 month periods since 1948 that were lower than the previous 12 months (as happened in 2025 compared to 2024). In Part 1 I did not have 12 months yet that were lower than the previous 12.

 

The table below shows when the 12 month average first fell and then the month when it stopped falling. The next column shows how many months in a row the 12 month average was lower than the previous 12 month average.

 

So, for example, the average for the 12 months ending in Dec. 1949 was lower than it was for Dec. 1948. Then the 12 month average for each of the next 9 months were also lower than the preceding 12 month average. 

 

The last column shows when the nearest recession occurred. All of these episodes overlap with a recession (some explanation of this after the table) except the current episode of 2025 being lower than 2024. If we don't have a recession soon it will mean that we are in a unique case. It could be the first time the 12 month average fell without a recession.

 

Start Date	End Date	Months	Closest Recession
Dec. 1949	Sept. 1950	10	Nov. 1948-Oct. 1949
Feb. 1954	Jun. 1955	17	Jul. 1953-May 1954
Feb. 1958	May. 1959	16	Aug. 1957-Apr. 1958
Feb. 1961	Apr. 1962	15	Apr. 1960-Feb. 1961
Oct. 1970	May. 1972	20	Dec. 1969-Nov. 1970
Feb. 1975	May. 1976	16	Nov. 1973-Mar. 1975
Oct. 1980	Jul. 1981	10	Jan. 1980-Ju. 1980
May. 1982	Nov. 1983	19	Jul. 1981-Nov. 1982
Nov. 1990	Aug. 1993	34	Jul. 1990-Mar. 1991
Feb. 2001	Sept. 2004	44	Mar. 2001-Nov. 2001
Jan. 2008	Oct. 2011	46	Dec. 2007-June 2009
Apr. 2020	Aug. 2021	19	Feb. 2020-Apr. 2020
Jan. 2026

 

The first row of this table tells us that the percentage was lower for all of 1949 than it was for all of 1948. The recession did not end until Oct. 1949. So that episode clearly overlaps with a recession.

 

Also notice that there is never any case of less than 10 months. So the odds are we will see more of this right now.  One thing that is different now is that we have tariffs going up and down and immigration has been reduced. Maybe those have had or will have some effect on this. 

See US Business Cycle Expansions and Contractions from The National Bureau of Economic Research.

Also see Employment-Population Ratio - 25-54 Yrs. from The Federal Reserve Bank of St. Louis.

See What's the Sahm Rule? Alarming Jobs Report Raises Recession Risk: A key indicator has predicted every recession since 1970, and the alarm just sounded by Eric Boehm of Reason (this was from Aug. 2024). Excerpt:

"It is named after economist Claudia Sahm, who served as a top economic advisor during the Obama administration and identified a historical indicator of coming recessions in 2019: every time since 1970 that the three-month moving average of the U.S. unemployment rate is more than half a percentage point above the lowest three-month moving average from the previous year, a recession has soon followed."