Unemployment Rate

The national unemployment rate, officially U-3, is unemployed people as a share of the labor force, all employed plus unemployed people. The Bureau of Labor Statistics (BLS), part of the U.S. Department of Labor, estimates it from the Current Population Survey (CPS).

The CPS is a monthly survey of about 60,000 households and about 110,000 individuals, chosen to represent every state plus the capital district. Census Bureau interviewers collect answers for BLS, which weights and publishes them monthly since 1940. It covers civilians age 16 and over, counts each person once, and excludes active-duty military and institutional residents, so it never counts unemployment-insurance claims directly.

Labor-force status always refers to the reference week, the Sunday-to-Saturday week including the 12th, with interviews in the following week. BLS releases The Employment Situation monthly, normally on the first Friday after the reference month at 8:30 a.m. Eastern Time under embargo. Each release also brings employer payroll estimates by industry, detailed household tables, and U-1 to U-6 gauges of wider labor underuse. BLS re-estimates seasonal factors every month for that month only, revises 5 years each January, and adds new population controls each January.

A monthly change of about 100,000 is significant in the employer survey, against about 500,000 in the household survey.

The rate has no fixed weights because it divides unemployed people by the labor force, so each breakdown shows who lost work.

• Reason for unemployment: job losers and temporary layoffs, quitters, reentrants, and new entrants; temporary layoffs need no job search.
• Duration of unemployment: people jobless for under 5 weeks, 5 to 14 weeks, 15 to 26 weeks, and 27 weeks or more.
• Demographics: adult men, adult women, teenagers, and White, Black, Asian, and Hispanic groups.
• Work-intensity margins: people working part time for economic reasons, plus marginally attached people who want work but stopped searching, including discouraged workers.
• Underutilization ladder, U-1 to U-6: from long-term jobless and job losers through U-3, then adding discouraged, marginally attached, and part-time workers for economic reasons.

April 2020, reported May 8, 2020, set records for level and monthly gain at 14.7%, up 10.3 points, as unemployed people rose 15.9M to 23.1M. Temporary layoffs drove it, up about ten-fold to 18.1M against only 2.0M permanent job losers. People jobless for under 5 weeks held about two-thirds (14.3M) while spells of 27 weeks or more fell to 939,000. The wider U-6 gauge reached 22.8% against 14.7% for U-3.

Labor-force participation fell to 60.2%, the lowest since January 1973, and the employment-population ratio hit a record low of 51.3%. Misclassified absent workers would have added almost 5 points on an unadjusted basis, though BLS kept the data as recorded.

1. Classify each person: Employed means any paid work, at least 15 hours of unpaid family work, or temporary absence. Unemployed means no job plus active search in 4 weeks and availability, or waiting on temporary recall; any work takes precedence over search.
2. Weight to the population: each answer is scaled to independent Census population totals, so 1 respondent stands for about 2,500 people. Weighted employed plus weighted unemployed people form the labor force.
3. Divide into rates: U-3 equals unemployed people divided by the labor force, times 100, so 23.1M divided by 156.5M gives about 14.7%. Participation divides the labor force by the population, and the employment share divides employed people by the population.
4. Adjust for seasons: strip predictable calendar effects using factors reestimated with data through that month, and publish the seasonally adjusted U-3 as the headline.