Last updated 3 weeks ago. Our resources are updated regularly but please keep in mind that links, programs, policies, and contact information do change.
The monthly unemployment rate comes from two massive surveys: one interviewing 60,000 American households and another querying hundreds of thousands of businesses. These surveys tell a richer story than any single percentage can capture.
The Official Unemployment Rate
The most cited unemployment figure results from a specific formula and strict definitions. The official unemployment rate represents unemployed people as a percentage of the total labor force.
The formula is: Unemployment Rate = (Number of Unemployed / Labor Force) × 100
The labor force includes all people currently employed plus all people currently unemployed. This specific measure carries the official designation “U-3” because it’s the third of six distinct measures of labor underutilization that the BLS publishes monthly.
Who Counts as What
The BLS classifies every person aged 16 and over in the “civilian noninstitutional population” into three groups: employed, unemployed, or not in the labor force. The civilian noninstitutional population excludes active-duty military personnel and people in institutions like prisons or nursing homes.
The classification follows a rigid hierarchy. If someone could fit multiple categories, having a job takes precedence over looking for one, and any labor force activity takes precedence over non-labor force activities like attending school.
The Employed
Someone counts as employed if they meet any of these criteria during the survey’s reference week:
- Did any work for pay or profit, including part-time and temporary work
- Worked at least 15 hours without pay in a family business or farm
- Had a job but were temporarily absent, whether paid or unpaid, for reasons like vacation, illness, childcare, maternity leave, bad weather, or labor disputes
This broad definition means a person who worked one paid hour during the week counts as employed, the same as someone working 40 hours with full benefits. The unemployment rate measures labor market activity, not economic hardship.
The Unemployed
To count as unemployed, someone must simultaneously meet three strict criteria during the reference week:
- Have no employment according to the definition above
- Be available for work
- Have actively searched for employment during the four-week period ending with the reference week
“Active efforts” includes contacting employers directly, interviewing, submitting resumes, filling out applications, placing or answering job ads, or using employment agencies. One exception: people on temporary layoff expecting recall don’t need to search actively to count as unemployed.
Not in the Labor Force
This category includes everyone who doesn’t meet the criteria for employed or unemployed. It covers full-time students, retirees, people caring for family members, and critically, jobless people who want work but aren’t actively searching. These “discouraged workers” or “marginally attached” workers don’t count in the official U-3 unemployment rate.
The Source: Current Population Survey
The unemployment rate comes from the Current Population Survey (CPS), a monthly survey of about 60,000 households across the United States. The Census Bureau has conducted this survey for the BLS since 1940.
Each month, interviewers ask detailed questions about household members’ work and job search activities. The questions refer to a specific “reference week”—the calendar week including the 12th day of the month. These responses determine how each person gets classified.
The survey uses a rotating sample where households participate for four months, rest for eight months, then return for another four months. This design ensures 75% of households remain the same month-to-month and 50% remain the same year-to-year, reducing volatility in estimates.
Two Surveys Tell the Full Story
The monthly “Jobs Report” actually combines two distinct surveys. The Current Population Survey provides the unemployment rate by surveying people. The Current Employment Statistics (CES) survey counts jobs on business payrolls. These surveys offer different but complementary views of the labor market.
The Household Survey: People’s Stories
The CPS captures rich demographic information because it surveys people, not businesses. The BLS can break down employment and unemployment data by age, sex, race, ethnicity, education, and marital status. The survey covers all worker types, including self-employed, agricultural workers, and unpaid family workers.
The survey’s rotating design provides statistical stability. With 75% of households remaining the same month-to-month, the survey can track changes in individual circumstances over time while maintaining representative sampling.
The Establishment Survey: Job Counts
The CES survey provides the other major headline number: the change in total nonfarm payroll employment. This massive undertaking surveys approximately 121,000 businesses and government agencies representing about 631,000 individual worksites.
Each month, these establishments report employees on their payroll who received pay during the pay period including the 12th of the month. The survey also collects data on average weekly hours and average hourly earnings.
The CES’s strength lies in detailed industry-level data. Using the North American Industry Classification System, the survey shows which economic sectors are adding or shedding jobs, from social assistance to federal government employment.
Why the Numbers Sometimes Differ
The CPS counts employed people while the CES counts payroll jobs. This fundamental difference explains apparent discrepancies:
Multiple Jobholders: Someone with two part-time jobs counts once as “employed” in the household survey but twice on different company payrolls in the establishment survey.
Coverage Scope: The CPS includes self-employed individuals, agricultural workers, and unpaid family workers. The CES focuses only on nonfarm payroll employment.
Statistical Reliability: The CES has a much larger sample size and uses administrative payroll records, making it more statistically robust for measuring month-to-month changes. A change of about 115,000 jobs is statistically significant for the CES, while the threshold for the CPS is around 600,000.
This reliability difference explains why economists focus on CES payroll numbers for monthly job growth while relying on the CPS for the unemployment rate. Only the CPS can calculate unemployment because it interviews people to determine if they’re jobless and actively seeking work.
| Feature | Current Population Survey (CPS) | Current Employment Statistics (CES) |
|---|---|---|
| Survey Type | Household Survey | Establishment (Payroll) Survey |
| What is Measured? | Labor force status of people | Number of nonfarm payroll jobs, hours, earnings |
| Source of Data | Interviews with ~60,000 households | Payroll records from ~121,000 businesses |
| Coverage | Total civilian population (includes self-employed, agricultural, unpaid family workers) | Private nonfarm and government payroll jobs only |
| Key Outputs | Official unemployment rate, labor force participation, demographic data | Change in nonfarm payrolls, average hourly earnings, average workweek |
| Statistical Volatility | Higher month-to-month volatility | Lower month-to-month volatility |
Local Unemployment: From National to Your City
While national unemployment rates get the most attention, state and local figures matter more for individual economic experiences. The BLS produces local figures through sophisticated modeling because the national survey lacks statistical power when divided among 50 states and thousands of metropolitan areas.
The Small Sample Problem
The 60,000-household national CPS becomes statistically weak when split geographically. Ohio might have only 2,000 households in the monthly sample, leading to large margins of error and volatile fluctuations that don’t reflect real economic changes. The BLS advises against using raw monthly CPS data for states due to reliability concerns.
The LAUS Program
The Local Area Unemployment Statistics (LAUS) program addresses this challenge through statistical modeling rather than additional surveys. LAUS generates monthly employment and unemployment estimates for over 7,500 areas nationwide.
For state estimates, LAUS uses a “signal-plus-noise” time series regression model. The model takes direct CPS estimates as the primary “signal” and incorporates CES payroll employment estimates and unemployment insurance claims data to reduce statistical “noise.”
All state estimates are “controlled”—mathematically adjusted so they sum precisely to national totals from the CPS. This ensures consistency across geographic levels.
Local Areas: The Handbook Method
For metropolitan areas, counties, and cities, the BLS uses the “Handbook Method,” building estimates from administrative and survey data:
Estimating the Unemployed: The model starts with people receiving unemployment benefits in that area, then adds statistical estimates for others not receiving benefits—people who exhausted benefits, new labor force entrants, and reentrants.
Estimating the Employed: The model takes job counts from the CES or Quarterly Census of Employment and Wages, adjusting for commuting patterns using Census Bureau data to estimate employed residents rather than just jobs located in the area.
Local estimates are controlled to sum to statewide totals, creating a hierarchical system from the smallest city to the national level. This ensures internal consistency but means any biases in national CPS data cascade through all geographic levels.
Weekly Unemployment Claims: A Different Measure
Every Thursday morning, the U.S. Department of Labor releases the Weekly Unemployment Insurance Claims report, often discussed alongside the monthly unemployment rate. This represents fundamentally different data—an administrative count rather than a statistical survey.
Administrative vs. Survey Data
The weekly UI claims report counts individuals who file for unemployment benefits through state programs. Unlike the CPS, this isn’t based on a survey sample but on actual administrative records from the unemployment insurance system.
Filing for benefits is an active, ongoing requirement. To receive payments, people must file claims for each week they remain unemployed, typically through online portals or automated phone systems. During weekly certification, claimants answer questions about availability for work, any earnings, and job search activities.
Initial vs. Continued Claims
The weekly report highlights two key figures:
Initial Claims: First-time filers for UI benefits after losing jobs. Economists watch this closely as a leading economic indicator because spikes in initial claims provide real-time signals of new layoffs, often preceding broader economic downturns.
Continued Claims (Insured Unemployment): People who filed initial claims and continue receiving benefits weekly. This measures unemployment persistence among the insured population and serves as a coincident or lagging economic indicator.
Why Claims Don’t Equal Unemployment
The number receiving unemployment benefits is consistently lower than total unemployed people measured by the CPS. UI claims represent a subset, not the complete unemployment picture:
Eligibility Requirements: Not everyone jobless qualifies for benefits. Applicants must have lost jobs through no fault of their own and have sufficient prior earnings history. This excludes new labor force entrants, long-term reentrants, and people who quit voluntarily.
Non-Participation: Some eligible individuals choose not to file claims for personal reasons.
Benefit Exhaustion: State benefits typically last up to 26 weeks. Once exhausted, people no longer appear in continued claims data even if they remain jobless and job-seeking.
Economists use weekly claims as a timely early warning system for trends that might later appear in comprehensive monthly reports. The data measures different things on different timelines, explaining why Thursday reports of rising “jobless claims” can precede Friday reports of stable “unemployment rates.”
The Complete Picture: U-1 Through U-6
The official U-3 unemployment rate provides consistent, objective measurement but doesn’t capture the full extent of joblessness or underemployment. The BLS calculates six “alternative measures of labor underutilization” monthly, designated U-1 through U-6, presenting a spectrum of labor market perspectives.
Why One Number Isn’t Enough
The official unemployment concept has been consistent since the 1940s. Its strength lies in objective definition: jobless, available for work, and actively searching. This precision comes at the cost of breadth, excluding people who want jobs but stopped looking and those employed but unable to get needed hours.
The BLS developed alternative measures, first introduced in 1976 and revised to the current U-1 through U-6 format in 1994. These measures acknowledge the inherent limitations of the U-3 rate and demonstrate institutional commitment to statistical transparency.
The Six Measures Explained
The “U” rates form a spectrum from most restrictive to most comprehensive unemployment definitions. While all six tend to move together over business cycles, the gaps between them reveal labor market quality, not just unemployment quantity.
U-1: Long-Term Unemployed The narrowest measure, focusing on severe joblessness. Includes only people unemployed 15 weeks or longer as a percentage of the civilian labor force.
U-2: Job Losers Focuses on involuntary unemployment. Includes individuals who lost jobs or whose temporary jobs ended as a percentage of the civilian labor force. Excludes voluntary job leavers and new/re-entrants to the labor force.
U-3: The Official Unemployment Rate The headline figure. Includes all unemployed people as a percentage of the civilian labor force.
U-4: Including Discouraged Workers The first broader measure. Adds discouraged workers to both the unemployed count and labor force. Discouraged workers are people not in the labor force who want jobs but stopped searching because they believe no jobs are available due to lack of work, previous inability to find jobs, or discrimination.
U-5: Including All Marginally Attached Workers Expands further by adding all marginally attached workers. These are people not currently in the labor force who want and are available for jobs and looked for work sometime in the past 12 months (but not the last four weeks). This includes discouraged workers plus those who stopped looking for other reasons like family responsibilities or school.
U-6: The Broadest Measure The most comprehensive labor underutilization measure. Takes the U-5 measure and adds people employed part-time for economic reasons—working less than 35 hours weekly but wanting and available for full-time work. They work part-time because their hours were cut or they couldn’t find full-time jobs.
| Measure | Description | Who is Added to U-3? |
|---|---|---|
| U-1 | Persons unemployed 15 weeks or longer, as percent of civilian labor force | N/A (narrower subset of U-3) |
| U-2 | Job losers and persons who completed temporary jobs, as percent of civilian labor force | N/A (narrower subset of U-3) |
| U-3 | (Official Rate) Total unemployed, as percent of civilian labor force | Baseline Measure |
| U-4 | Total unemployed plus discouraged workers, as percent of civilian labor force plus discouraged workers | + Discouraged Workers |
| U-5 | Total unemployed plus all marginally attached workers, as percent of civilian labor force plus all marginally attached workers | + All Marginally Attached Workers |
| U-6 | Total unemployed, plus all marginally attached workers, plus total employed part time for economic reasons, as percent of civilian labor force plus all marginally attached workers | + All Marginally Attached Workers + Involuntary Part-Time Workers |
The Monthly Jobs Report and Its Limits
The culmination of this data collection occurs monthly on “Jobs Day” when the BLS releases its Employment Situation summary. This report moves financial markets worldwide but requires careful interpretation to understand its limitations and statistical adjustments.
Jobs Day: The First Friday
The Employment Situation report typically releases at 8:30 AM Eastern Time on the first Friday of every month. This major economic event synthesizes findings from both pillar surveys: the CPS provides unemployment rates and demographic data while the CES provides payroll employment changes, earnings, and workweek data.
The release usually includes a public statement from the U.S. Secretary of Labor offering the current administration’s interpretation of the data.
Common Critiques of the Official Rate
The official unemployment rate’s consistent, objective definition—unchanged since the 1940s—creates significant limitations in capturing full labor market experiences.
The Discouraged Worker Problem
The U-3 rate only counts active job seekers. When jobless individuals become frustrated and stop searching, they move from “unemployed” to “not in the labor force.” This creates misleading outcomes: during severe recessions, the official unemployment rate can decline because people give up hope and drop out entirely rather than finding jobs.
Conversely, early in economic recoveries, unemployment rates may rise as optimism returns and discouraged workers re-enter the labor force and resume searching, moving from “not in the labor force” back to “unemployed.” This makes the U-3 rate a poor indicator of labor market health during business cycle turning points.
The Underemployment Issue
Official statistics ignore work quality and quantity. Someone working one hour weekly at minimum wage counts as “employed” the same as someone working 40 hours in a high-skilled, salaried position. This approach misses major underemployment forms:
Involuntary Part-Time Work: The system doesn’t distinguish between voluntary and forced part-time work due to hour cuts or inability to find full-time positions.
Skills Mismatch: It ignores overqualified individuals, like master’s degree engineers working as baristas. While counted as employed, their skills are underutilized and earnings likely far below potential.
The U-6 rate attempts to quantify involuntary part-time work. A high U-6 rate despite low U-3 suggests an economy creating jobs but not necessarily high-quality, full-time employment.
Understanding Revisions and Adjustments
The BLS employs statistical procedures that can alter numbers after initial release. These aren’t errors but standard practices maintaining data integrity.
Seasonal Adjustment
Economic data often shows predictable seasonal patterns—retail employment swells during winter holidays, construction activity increases in summer. These regular fluctuations can obscure underlying, non-seasonal economic trends.
Seasonal adjustment is a statistical technique measuring and removing predictable seasonal influences from data. This process, using sophisticated programs like X-12-ARIMA, enables meaningful month-to-month comparisons. Most headline Jobs Report figures are seasonally adjusted, though non-adjusted data remains available for analysts studying seasonal patterns.
Data Revisions
The first CES job growth estimate released on Jobs Day is preliminary, based on survey responses received by that point. The BLS revises initial estimates in the following two months as more business data becomes available.
Additionally, the BLS performs annual “benchmark revisions,” adjusting CES estimates to align with more comprehensive Quarterly Census of Employment and Wages data based on state unemployment insurance tax records. Population controls for the CPS are simultaneously updated with new Census Bureau data. This annual process ensures long-term accuracy and reliability of the entire labor statistics suite.
Reading Between the Numbers
Understanding U.S. unemployment measurement requires recognizing both the sophistication and limitations of the system. The monthly unemployment rate represents one of the most comprehensive and consistent labor market measurements globally, yet it captures only part of the employment story.
The existence of six different unemployment measures (U-1 through U-6) acknowledges that no single number can fully represent labor market complexity. During economic expansions, gaps between these measures often narrow as both unemployment and underemployment decrease. During recessions, the gaps widen dramatically, revealing the full scope of labor market distress.
The dual-survey system—households and establishments—provides crucial cross-validation. When both surveys show similar trends, confidence in the direction of change increases. When they diverge, economists examine the differences to understand what’s happening beneath the surface.
Local unemployment statistics, while less precise than national figures, provide essential insights into regional economic conditions. The modeling techniques used to generate these estimates represent practical solutions to statistical limitations, acknowledging that perfect measurement is impossible while striving for useful approximation.
The weekly unemployment insurance claims data offers real-time labor market pulse checks, but users must remember its narrow scope—it measures only a subset of unemployed individuals and reflects the mechanics of the benefit system as much as underlying economic conditions.
Perhaps most importantly, the unemployment rate measures labor market activity, not economic hardship or well-being. A low unemployment rate doesn’t necessarily indicate broad prosperity, just as a high rate doesn’t automatically signal widespread suffering. The rate must be interpreted alongside other economic indicators—wages, productivity, inflation, labor force participation—to understand the full economic picture.
The monthly Jobs Report represents a remarkable achievement in economic measurement, providing timely, comprehensive, and largely accurate snapshots of a $25 trillion economy’s labor market. Understanding its methodology, limitations, and nuances is essential for anyone seeking to make sense of economic news and policy debates.
This measurement system, built over decades and refined continuously, reflects both the ambition and constraints of statistical science applied to human economic behavior. The unemployment rate will continue evolving as the economy changes, but its fundamental challenge remains constant: reducing the complexity of millions of individual work experiences into numbers that inform public understanding and policy decisions.
Our articles make government information more accessible. Please consult a qualified professional for financial, legal, or health advice specific to your circumstances.
