Labor force participation and employment: demographic detail
How much of the change in labor force participation and employment reflects an aging population rather than a change in behavior within age groups. The top charts show each rate’s actual path against an age/sex-adjusted path: each month’s change is split into a behavior part (people in each age-and-sex group changing what they do) and a composition part (the mix of groups shifting), only the behavior part is accumulated, and the result is anchored to the latest actual rate — so the adjusted line keeps 2020’s collapse, which was behavior, not aging. The peak selector below shows the opposite experiment, the aging-only hypothetical: each group’s behavior frozen at a chosen business-cycle peak, with only the population’s age/sex mix allowed to change. A separate bank at the bottom breaks the current level and change down by sex, age, race/ethnicity, and education, including arbitrary subgroup cross-cuts.
Actual vs. Age/Sex-Adjusted
Labor force participation rate
Employment-population ratio
Demographic contributions to the change
Aging-Only Counterfactual, Since a Chosen Peak
Labor Market by Group
CPS microdata (IPUMS)The left block is every dimension on its own. Set any of the four dropdowns to a specific value to restrict the right block to that subgroup — whatever's left on “Overall” stays broken out there, lined up row-for-row with the left block (a fixed dimension leaves its usual rows blank on the right rather than shifting anything below it). All on Overall = the two blocks match exactly.
Source: BLS Current Population Survey and CPS microdata (IPUMS); author's calculations. Age/sex-adjusted series (top): the behavior part of each month's change, accumulated and anchored to the latest actual rate (a chained calculation, so the history re-levels slightly each month). Aging-only counterfactuals (middle): each group's rate held at its value at the chosen peak, with only the population's age/sex composition evolving. The by-group tables below are raw (unadjusted) rates — differences across groups reflect both behavior and composition, and refresh on IPUMS's monthly schedule rather than jobs day. Small-sample cells are flagged with †.