Unemployment is 4.1%, below its
Is unemployment rising?
In brief
Is unemployment rising, and do the two job surveys agree?- Unemployment is 4.1%, down 0.4pp since Nov 2025 and below the .→ Exhibit 1
- Prime-age employment is 80.7%, up 0.1pp since Nov 2025.→ Exhibit
- Households report 989k fewer jobs than payrolls since Nov 2025; January's population update more than accounts for it.→ Exhibit
- The January 2026 population update cut December 2025 employment by 1.43 million, not a loss of jobs.→ Exhibit
| What would change this view | Trigger | Today | Decided by |
|---|---|---|---|
| Unemployment crosses the | U-3 three-month average > 4.63% (0.50pp above its twelve-month low) | 4.13% (Oct 2026)not crossed | Nov 2026 jobs report (BLS) |
| Prime-age employment falls | 25–54 employment rate < 80.4% (one 90% confidence interval away) | 80.7% (Oct 2026)not crossed | Nov 2026 jobs report (BLS) |
The reasoning behind each point
- SlackUnemployment is 4.1%, down 0.4pp since Nov 2025 and below the 2015–19 average.
Oct 2026. U-6, the broad measure that adds discouraged and involuntary part-time workers, is 7.9%. The three-month average is ±0.00pp from its low of the prior twelve months (the Sahm gauge; +0.50pp has marked recessions). U-3 compares with a 2015–19 average of 4.4% and sits at the 19th percentile since 1990; U-6 moved −0.8pp since Nov 2025 and part-time work for economic reasons −647k.
- Prime agePrime-age employment is 80.7%, up 0.1pp since Nov 2025.
The share of 25–54 year-olds in work. The prime-age ratio is not moved by ageing, and the January 2026 population controls moved it by −0.1pp against −0.5pp for the 16+ ratio (BLS), so it is the cleaner read of employment. Its high since 2000 is 81.9% (Apr 2000). Prime-age participation 83.7% (−0.1pp); total participation 61.9%, −0.1pp on the year on BLS's population-consistent basis against −0.6pp as published.
- Two surveysHouseholds report 989k fewer jobs than payrolls since Nov 2025; January's population update more than accounts for it.
Households report −145k jobs and payrolls +844k. The January 2026 population controls account for −1,385k of it (140%); on BLS's population-consistent series household employment changed +1,240k. On the payroll concept (excluding the self-employed, farm and private-household workers and unpaid absences, adding second jobs) the change is +1,363k, a concept difference of −123k, leaving +519k of disagreement between the surveys (sampling error in both, and genuine differences in coverage).
- PopulationThe January 2026 population update cut December 2025 employment by 1.43 million, not a loss of jobs.
It also lowered participation by 0.4pp. Population fell −231k but its mix shifted toward groups with lower participation; unemployment was barely changed (+15k). A year earlier the January 2025 controls added +2,871k people and +2,000k employed. Official levels are not revised for these breaks; ratios (unemployment rate, prime-age employment rate) are the comparable measures across Januaries.
Next: Claims — the flow into unemployment
Unemployment, and the slack behind it
Is unemployment rising, and how much slack is there?Exhibit 1 · Unemployment rates over time: headline (U-3), broad (U-6) and prime-age
About this exhibit
Unemployment's three-month average minus its low of the prior twelve months
About this exhibit
Unemployment and participation rates by measure, age and sex
| Series | Level | 1m chg | 3m chg | 12m chg | Level pct since 1990 | Level, 36 months | As of |
|---|---|---|---|---|---|---|---|
| Unemployment | |||||||
| Unemployment rate (U-3) | 4.1% | −0.1pp | ±0.0pp | −0.4pp | 19 | Oct 2026 | |
| U-6: plus marginally attached and part time for economic reasons | 7.9% | +0.3pp | ±0.0pp | −0.8pp | 27 | Oct 2026 | |
| U-1: unemployed 15 weeks or longer, % of labour force | 1.9% | +0.1pp | +0.2pp | +0.1pp | 57 | Oct 2026 | |
| Unemployment rate, 25–54 | 3.6% | ±0.0pp | ±0.0pp | −0.3pp | 29 | Oct 2026 | |
| Unemployment rate, men 20+ | 3.9% | ±0.0pp | ±0.0pp | −0.2pp | 29 | Oct 2026 | |
| Unemployment rate, women 20+ | 3.6% | ±0.0pp | −0.1pp | −0.5pp | 16 | Oct 2026 | |
| Unemployment rate, 16–19 | 13.9% | −0.6pp | +1.8pp | −2.4pp | 26 | Oct 2026 | |
| U-3 three-month average minus its prior 12-month lowcalculated by the site (Sahm-rule construction) | 0.0pp | ±0.0pp | – | – | 38 | Oct 2026 | |
| Participation | |||||||
| Participation rate, 16+ | 61.9% | +0.1pp | +0.5pp | −0.6pp | 6 | Oct 2026 | |
| Participation rate, 16–19 | 36.7% | −0.3pp | +1.8pp | +0.4pp | 41 | Oct 2026 | |
| Participation rate, 20–24 | 71.9% | +0.4pp | +1.7pp | +0.1pp | 42 | Oct 2026 | |
| Participation rate, 25–54 | 83.7% | ±0.0pp | +0.3pp | −0.1pp | 84 | Oct 2026 | |
About this exhibit
Prime-age employment rate since 2010
About this exhibit
Participation and employment rates as published and adjusted for population updates
About this exhibit
Household and payroll
How much of the gap between the two surveys is population arithmetic?Twelve-month employment change in the household survey against the payroll survey
About this exhibit
Household minus payroll job change over twelve months, split by term
About this exhibit
Household–payroll job gap by term, and household jobs converted to payroll basis
| Household survey, official | −145k |
| Payroll survey | +844k |
| Gap (household − payroll) | −989k |
| of which | |
| Population controlsofficial − population-consistent household change | −1,385k |
| Concepthousehold − payroll-concept household, both population-consistent | −123k |
| Residualpayroll-concept household − payroll survey | +519k |
| Household employment → payroll concept, not seasonally adjusted | Oct 2026 | 12m chg |
|---|---|---|
| Household survey employment | 164,183 | +117 |
| − Agriculture and related industries | 2,537 | −389 |
| − Nonagricultural self-employed, unincorporated | 8,832 | +667 |
| − Nonagricultural unpaid family workers | 69 | −1 |
| − Private household wage and salary workers | 614 | −30 |
| − Nonagricultural wage and salary workers with unpaid absences | 1,806 | −126 |
| + Nonagricultural wage and salary multiple jobholders | 8,672 | −348 |
| = Payroll concept, not seasonally adjusted | 158,997 | −110 |
Official levels, so the 12-month changes include the January 2026 population-control break. BLS seasonally adjusts the result to publish LNS16000000 (+109k over the same twelve months, official basis).
About this exhibit
Every level: employed, unemployed and outside the labour force
Every level in thousands: who is employed, unemployed and outside the labour force
Effect of each January's population update on levels and rates
| January | Population | Labour force | Employed | Unemployed | LFPR | EPOP 25–54 | U-3 |
|---|---|---|---|---|---|---|---|
| 2026 | −231 | −1,417 | −1,432 | +15 | −0.4 | −0.1 | ±0.0 |
| 2025 | +2,871 | +2,106 | +2,000 | +105 | +0.1 | ±0.0 | +0.1 |
| 2024 | −625 | −299 | −270 | −28 | ±0.0 | ±0.0 | ±0.0 |
| 2023 | +954 | +871 | +810 | +60 | +0.1 | ±0.0 | ±0.0 |
| 2022 | +973 | +1,530 | +1,471 | +59 | +0.3 | +0.1 | ±0.0 |
| 2021 | −476 | −200 | −180 | −20 | ±0.0 | ±0.0 | ±0.0 |
| 2020 | −811 | −524 | −507 | −17 | ±0.0 | ±0.0 | ±0.0 |
| 2019 | −800 | −506 | −488 | −18 | ±0.0 | ±0.0 | ±0.0 |
| 2018 | +488 | +333 | +318 | +15 | ±0.0 | ±0.0 | ±0.0 |
| 2017 | −831 | −508 | −487 | −21 | ±0.0 | ±0.0 | ±0.0 |
| 2016 | +265 | +218 | +206 | +12 | ±0.0 | ±0.0 | ±0.0 |
About this exhibit
Levels and changes of population, the labour force and unemployment
| Series | Level | 1m chg | 3m chg | 12m chg | YoY | 12m chg, pop.-consistent | Level, 36 months | As of |
|---|---|---|---|---|---|---|---|---|
| Levels | ||||||||
| Civilian noninstitutional population, 16+not seasonally adjusted | 275,682 | +128k | +400k | +1,049k | +0.4% | +1,199k | Oct 2026 | |
| Civilian labour force | 170,683 | +421k | +1,589k | −858k | −0.5% | +515k | Oct 2026 | |
| Employed | 163,615 | +463k | +1,438k | −145k | −0.1% | +1,240k | Oct 2026 | |
| Unemployed | 7,068 | −41k | +152k | −713k | −9.2% | −733k | Oct 2026 | |
| Not in the labour force | 104,999 | −293k | −1,190k | +1,907k | +1.9% | +702k | Oct 2026 | |
| Not in the labour force, want a job now | 6,024 | +234k | +104k | −115k | −1.9% | – | Oct 2026 | |
| Part time for economic reasons | 4,840 | +339k | +36k | −647k | −11.8% | – | Oct 2026 | |
| Unemployed 27 weeks or longer | 1,963 | +19k | +192k | +53k | +2.8% | – | Oct 2026 | |
| Unemployed: job losers | 3,166 | −34k | −143k | −394k | −11.1% | – | Oct 2026 | |
| Unemployed: job losers on temporary layoff | 766 | −79k | −155k | −236k | −23.6% | – | Oct 2026 | |
| Unemployed: job leavers | 766 | +25k | −27k | −95k | −11.0% | – | Oct 2026 | |
| Unemployed: re-entrants | 2,297 | +8k | +174k | −303k | −11.7% | – | Oct 2026 | |
About this exhibit
Appendix
Survey, , concepts and the October 2025 gap15 checks, all pass
The household survey measures people rather than jobs: who is working, looking, or out of the labour force. Its levels jump every January when the survey is re-weighted to new population estimates, so this page reads slack from ratios, shows BLS's population-consistent series beside the official ones, and splits the gap with the payroll survey into population arithmetic, concepts and residual.
Definitions and caveats
- SurveyPeople, not jobs.
The Current Population Survey is a sample of about 60,000 eligible households a month (BLS Technical Note), asked about the week that includes the 12th. It counts people (a multiple jobholder once), includes the self-employed, farm and private-household workers, and people on unpaid leave. The Technical Note puts the 90% confidence interval on the monthly change in unemployment at about ±425,000 and in the unemployment rate at about ±0.3 percentage point (at an unemployment rate near 6%).
- PopulationA break every January.
Each January (February in 2026, after the shutdown) the survey is re-weighted to new Census Bureau population estimates. Official levels are not revised backward, so the December–January change in levels carries the whole cumulative correction since the 2020 Census. BLS publishes the effect on the prior December (table above); in 2025 and 2026 it also published experimental series that spread the effect over April 2020 onward.
- ConsistentBLS experimental series, then official.
The population-consistent basis uses BLS's experimental series (April 2020–December 2025, accounting for the January 2026 controls) and official estimates from January 2026, which already carry the new controls. BLS advises against combining the experimental series with official data before April 2020, so the basis starts there.
- ConceptHousehold employment on the payroll concept.
The household survey adjusted to the payroll concept is a BLS research series (LNS16000000): household employment minus agriculture, the unincorporated self-employed, unpaid family workers, private household workers and wage and salary workers on unpaid absences, plus multiple jobholders' second jobs, then seasonally adjusted.
- GapThree terms that add up.
Household minus payroll employment change over twelve months = (official − population-consistent household change) + (population-consistent household − population-consistent payroll-concept household change) + (payroll-concept household − payroll change). The first term is the population arithmetic, the second the difference in concepts, the third sampling error and genuine disagreement.
- Sahm ruleThree-month average against the low.
Indicator computed here on BLS data: the three-month average of the unemployment rate minus its lowest three-month average over the prior twelve months (Claudia Sahm's construction). The October 2025 gap is bridged by averaging the available months in each three-month window.
- Data gapOct 2025: not collected.
October 2025 household survey not collected (federal government shutdown; BLS footnote 9). Left empty: no interpolation. A comparison with a year earlier that lands on it compares with the next month collected (November 2025), and the text says "since" that month (rule of 8 Nov 2026).
- NormsOwn history, 2015–19.
Shading compares each change with its own 2015–19 distribution; percentiles of the level run from January 1990.
Sources
- publicU.S. Bureau of Labor Statistics, Current Population Survey (CPS) labor force statistics, flat file ln.data.1.AllData
Source: U.S. Bureau of Labor Statistics · https://download.bls.gov/pub/time.series/ln/
- publicU.S. Bureau of Labor Statistics, The Employment Situation — August 2026, Table A-1 (USDL-26-1435)
Source: U.S. Bureau of Labor Statistics · https://www.bls.gov/news.release/empsit.t01.htm
- publicU.S. Bureau of Labor Statistics, CPS population control adjustment effects, key series, January 2016–2026 (XLSX) and Adjustments to household survey population estimates in January 2026
Source: U.S. Bureau of Labor Statistics · https://www.bls.gov/cps/documentation.htm#pop
- publicU.S. Bureau of Labor Statistics, Experimental series accounting for January 2026 population control effects, April 2020–December 2025 (XLSX)
Source: U.S. Bureau of Labor Statistics (experimental series) · https://www.bls.gov/cps/methods/population-controls/experimental-series-accounting-for-january-2026-population-control-effects.htm
- publicU.S. Bureau of Labor Statistics, Current Employment Statistics (total nonfarm, for the comparison)
Source: U.S. Bureau of Labor Statistics · https://download.bls.gov/pub/time.series/ce/
Reproduction checks
| Check | Published | Computed | Tolerance | Result |
|---|---|---|---|---|
| Table A-1 rows mapped to CPS series (49 rows in 6 groups; unmapped) | 0 | 0 | 0 | pass |
| Table A-1, Oct 2026: 343 published values (levels, rates, seasonally adjusted and not, 9 months) reproduced from the flat file (values that differ) | 0 | 0 | 0 | pass |
| Unemployment rate Oct 2026 vs Table A-1 (%) | 4.1 | 4.1 | 0.051 | pass |
| U-3 = unemployed ÷ labour force within the rounding bound of the published inputs (0.05pp + level rounding), 6 groups × every month since Jan 1990 (2646 observations; violations) | 0 | 0 | 0 | pass |
| Employment–population ratio = employed ÷ population within rounding, same 2646 observations (violations) | 0 | 0 | 0 | pass |
| Participation rate = labour force ÷ population within rounding, same 2646 observations (violations) | 0 | 0 | 0 | pass |
| Labour force = employed + unemployed within 1k, same 2646 observations (violations) | 0 | 0 | 0 | pass |
| Not in labour force = population − labour force within 1k, same 2646 observations (violations) | 0 | 0 | 0 | pass |
| BLS experimental workbook, 'Official series' tab (Apr 2020–Jan 2026, 9 series, 621 values) equals the CPS flat file (values that differ) | 0 | 0 | 0 | pass |
| Population-control workbooks 2020–2026: December values 'as published' equal the not seasonally adjusted flat file, and 'difference' = updated − published (112 values; values that differ) | 0 | 0 | 0 | pass |
| January 2026 population-control workbook: December 2025 values 'based on updated population controls' equal the experimental series (not seasonally adjusted, 8 measures; values that differ) | 0 | 0 | 0 | pass |
| January 2026 population-control effect on employment equals official − experimental, December 2025, not seasonally adjusted (thousands) | -1432 | -1432 | 1.01 | pass |
| CPS headline series missing only in documented months since 1990 (2025-10; undocumented missing months) | 0 | 0 | 0 | pass |
| Household–payroll 12-month gap = population controls + concept + residual (identity, thousands) | -989 | -989 | 0.001 | pass |
| Bridge table: the published 12-month changes of the rows sum to the published change of the payroll concept (thousands) | -110 | -110 | 0.001 | pass |
- U.S. Bureau of Labor Statistics, Current Population Survey (CPS) labor force statistics, flat file ln.data.1.AllData — Source: U.S. Bureau of Labor Statistics
- U.S. Bureau of Labor Statistics, The Employment Situation — August 2026, Table A-1 (USDL-26-1435) — Source: U.S. Bureau of Labor Statistics
- U.S. Bureau of Labor Statistics, CPS population control adjustment effects, key series, January 2016–2026 (XLSX) and Adjustments to household survey population estimates in January 2026 — Source: U.S. Bureau of Labor Statistics
- U.S. Bureau of Labor Statistics, Experimental series accounting for January 2026 population control effects, April 2020–December 2025 (XLSX) — Source: U.S. Bureau of Labor Statistics (experimental series)
- U.S. Bureau of Labor Statistics, Current Employment Statistics (total nonfarm, for the comparison) — Source: U.S. Bureau of Labor Statistics