How the Government Counts Unemployment Each Month
What the unemployment rate actually measures
The unemployment rate is not a count of everyone without a job. It is the percentage of people actively looking for work who cannot find it. The U.S. Bureau of Labor Statistics (BLS) calculates it by dividing the number of unemployed people by the total labor force — which includes only people who are working or actively searching for work.
This means the rate excludes people who have stopped looking, retired people, full-time students not seeking work, and people unable to work. A person must be jobless, able to work, and actively searching (through job applications, interviews, or agency contact) to count as unemployed. Someone who quit their job but is not yet looking does not count. Someone who gave up searching last month does not count either.
The distinction matters because the unemployment rate can stay flat or even drop while more people are out of work — if enough people stop searching, the denominator shrinks and the percentage can improve on paper while the actual situation worsens.
Key Takeaways
- The unemployment rate is calculated by dividing the number of unemployed people by the total labor force, not by the total population.
- To be counted as unemployed, a person must be without work, able to work, and actively searching through applications, interviews, or agency contact in the past four weeks.
- The BLS collects this data monthly through the Current Population Survey, which interviews about 60,000 households across the United States.
- The unemployment rate can appear to improve even when more people are jobless if enough people stop searching and leave the labor force.
- Different unemployment measures (U-3 through U-6) exist because the official rate does not capture underemployment or discouraged workers.
The Current Population Survey: where the numbers come from
Every month, the BLS conducts the Current Population Survey (CPS), a survey of roughly 60,000 households across all 50 states. Census Bureau staff call or visit these households and ask detailed questions about employment status, job search activity, and work history. The survey is designed to represent the entire U.S. civilian population age 16 and older.
The same households are surveyed for four consecutive months, then rotated out and replaced. This overlap allows the BLS to track month-to-month changes and spot trends. The survey happens in the week that includes the 12th of each month, and results are released on the first Friday of the following month.
Because it is a sample rather than a complete count, the CPS has a margin of error. The BLS publishes confidence intervals alongside the headline rate, though these are rarely reported in news coverage. A change of 0.1 or 0.2 percentage points in the unemployment rate can fall within the margin of error and may not represent a real shift.
How the BLS defines "actively searching"
The BLS uses a specific definition of job search to separate the unemployed from the out-of-work population. In the four weeks before the survey, a person must have taken at least one of these actions: submitted a job process, interviewed with an employer, contacted a public or private employment agency, contacted friends or relatives about a job, placed or answered a job advertisement, or attended a job training program.
Passive activities do not count. Reading job postings online, updating a resume, or thinking about looking for work are not enough. The person must have taken a concrete step to contact an employer or agency. This rule is why the unemployment rate can seem disconnected from the number of people struggling to find work — many people who are out of work but have not taken a formal step in the past month are not counted.
If someone was laid off and is waiting to be recalled to their previous job, they are counted as unemployed even if they have not actively searched. The BLS recognizes that some people are genuinely attached to the labor market even if they are temporarily inactive.
The difference between U-3 and other unemployment measures
The headline unemployment rate reported in the news is called U-3. It is the official measure and the one most commonly cited. But the BLS also publishes five other unemployment measures — U-1 through U-6 — that paint a broader picture of joblessness and underemployment.
U-1 counts only people unemployed for 15 weeks or longer. U-2 counts people who lost a job or completed a temporary assignment. U-4 adds discouraged workers — people who have stopped searching because they believe no jobs are available for them. U-5 adds other marginally attached workers — people who want work and are available but have not searched in the past month. U-6, the broadest measure, includes all of U-5 plus people working part-time involuntarily because they cannot find full-time work.
U-6 is typically two to three percentage points higher than U-3 in normal economic times, and the gap widens during recessions. For example, if the headline rate is 4 percent, U-6 might be 7 percent. The BLS publishes all six measures monthly, but only U-3 gets mainstream media attention.
Why the unemployment rate can mislead
The unemployment rate is useful for tracking broad economic trends, but it has blind spots. It does not capture underemployment — someone working 10 hours a week at minimum wage while needing full-time work counts as employed. It does not capture wage decline or job quality. It does not reflect how long people have been searching or how many have given up.
During economic downturns, the unemployment rate can understate hardship because discouraged workers leave the labor force. In 2009, the headline unemployment rate peaked at 10 percent, but U-6 reached 17.1 percent. The difference represents millions of people who were out of work but no longer actively searching.
The rate also varies significantly by demographic group. Unemployment rates for Black workers, Hispanic workers, and young workers are typically higher than the national average, sometimes by two to three percentage points. A single national rate masks these disparities.
How state and local unemployment rates are calculated
The BLS also publishes unemployment rates for each state and for major metropolitan areas. State rates are calculated using the same CPS data, but with smaller sample sizes, which means larger margins of error. Some states are oversampled to improve reliability, while others rely on smaller samples.
In addition to the CPS, the BLS uses unemployment insurance claims data to estimate state and local rates. When someone files for unemployment benefits, that claim is reported to the state labor department and then to the BLS. This administrative data provides a real-time count of people receiving benefits, though it does not measure unemployment directly — it measures benefit receipt, which is narrower.
State unemployment rates are released monthly, usually one week after the national rate. Local area unemployment rates (for counties and metropolitan areas) are released quarterly or annually, with a longer lag, because the sample sizes are even smaller.
What happens when the labor force shrinks or grows
The unemployment rate depends on both the numerator (number of unemployed) and the denominator (size of the labor force). When the labor force shrinks, the unemployment rate can fall even if the number of jobless people stays the same or rises. This happened in 2020 during the COVID-19 pandemic: the unemployment rate fell from 14.7 percent in April to 6.7 percent by December, but millions of people had left the labor force and were not counted as unemployed.
The labor force grows when population increases, when more people enter the job market (such as younger cohorts reaching working age), or when labor force participation rises (more people choosing to work or search for work). It shrinks when people retire, when participation falls (people leaving the workforce), or when population declines.
The BLS tracks labor force participation rate separately from the unemployment rate. This rate shows what percentage of the civilian population age 16 and older is either working or actively searching. A falling participation rate signals that people are leaving the labor force, which can happen during recessions, during demographic shifts, or when people become discouraged.
Frequently Asked Questions
Why does the unemployment rate sometimes go down when jobs are lost?
The rate can fall if more people stop searching than lose jobs, because the labor force shrinks. If 100,000 people lose jobs but 150,000 leave the labor force, the unemployment rate can decline even though the total number of jobless people rose. This is why the BLS publishes labor force participation data alongside the unemployment rate.
How long do you have to be out of work to count as unemployed?
There is no minimum duration. You count as unemployed the moment you lose your job and take a concrete step to search for work within the past four weeks. You remain unemployed as long as you are jobless and actively searching. Once you stop searching for four weeks, you drop out of the labor force and are no longer counted as unemployed.
Does the unemployment rate include people on unemployment benefits?
Not automatically. You must meet the BLS definition of unemployed — jobless, able to work, and actively searching — to be counted, regardless of whether you receive benefits. Some people on benefits are not actively searching and are not counted. Some people actively searching are not may be able to access for benefits and are still counted as unemployed.
What is the difference between the unemployment rate and the number of people on unemployment?
The unemployment rate is a percentage of the labor force calculated from survey data. The number of people receiving unemployment benefits is an administrative count from state labor departments. The two numbers measure different things and rarely match — the unemployment rate is typically higher because it includes people not receiving benefits.
How accurate is the unemployment rate?
The CPS has a margin of error of roughly plus or minus 0.2 percentage points at the 90 percent confidence level. This means a reported rate of 4.0 percent could actually be anywhere from 3.8 to 4.2 percent. Month-to-month changes smaller than this margin may not represent real economic shifts, though the BLS does not always flag this in its releases.
This guide is for general information only and is not medical, financial, legal, or other professional advice. For decisions specific to your situation, consult a qualified professional. See our Editorial Policy.