How Unemployment Is Measured: The Unemployment Rate Explained

The unemployment rate is the share of the labor force that is jobless, available for work and actively looking for it. In the United States it is produced by the Bureau of Labor Statistics (BLS) from a monthly household survey, not from a count of people receiving unemployment benefits. That distinction explains a good deal of what confuses readers about the monthly jobs report.

The headline rate is a ratio of two groups. Dividing the number of unemployed people by the labor force gives a percentage that can move for reasons unrelated to whether jobs are being created, because people can enter or leave the labor force. Understanding who is counted, and who is not, is the key to reading the number.

In brief

• The unemployment rate comes from the Current Population Survey (CPS), a monthly survey of households.
• To be counted as unemployed, a person must have no job, be available to work and have searched for work in the previous four weeks, with an exception for people waiting to be recalled from layoff.
• The rate is unemployed divided by the labor force, which is the employed plus the unemployed.
• Nonfarm payrolls come from a different survey, of employers, which is why the two sets of numbers can disagree.
• The BLS also publishes broader measures of labor underutilization.

The survey behind the number

The CPS is conducted by the Census Bureau for the BLS. According to the BLS, trained interviewers contact about 60,000 eligible households each month, covering roughly 110,000 people, in person or by phone. The sample rotates, so that about 75 percent of households stay in the sample from one month to the next. Responses are weighted to Census population estimates by age, sex, race, Hispanic ethnicity and state, so that the results represent the civilian noninstitutional population aged 16 and older.

The survey asks about activity during a specific reference week, the week that includes the 12th of the month. Interviewers do not ask people whether they consider themselves unemployed. They ask a series of questions about work, job search and availability, and the BLS classifies each person from the answers. That approach keeps the definition consistent and comparable over time.

Who counts as what

The CPS sorts the population aged 16 and over into three categories.

Employed. Anyone who did any paid work during the reference week, or worked at least 15 hours without pay in a family business or farm. People who have a job but were temporarily absent, because of vacation, illness or a labor dispute, are also counted as employed.

Unemployed. People who had no job during the reference week, were available for work and made specific efforts to find a job in the previous four weeks. People who are waiting to be recalled from a temporary layoff are counted as unemployed even if they did not search.

Not in the labor force. Everyone else: students, retirees, caregivers, people with disabilities and those who are not looking for work. Within this group, the BLS identifies people who are marginally attached, meaning they want a job and have looked for one within the past 12 months but did not search in the last four weeks. Discouraged workers are a subset of that group who say they stopped looking because they believe no jobs are available.

The labor force is the employed plus the unemployed. The unemployment rate is the number of unemployed divided by the labor force. The labor force participation rate is the labor force divided by the civilian noninstitutional population aged 16 and older. The employment-population ratio is the number employed divided by that same population.

Why the definition produces surprises

The most common misunderstanding is that the rate counts everyone without a job. It does not. A person who has stopped looking is no longer counted as unemployed. When people leave the labor force, the unemployment rate can fall even if no new jobs were created. In the same way, when people who had not been looking begin a search, the rate can rise even if the economy is adding jobs, because they now count as unemployed.

That is why economists read the unemployment rate alongside participation and the employment-population ratio. A falling rate with a falling participation rate tells a different story from a falling rate with rising employment.

Another common assumption is that the rate is based on unemployment insurance claims. It is not. Many unemployed people never apply for benefits, and many are ineligible. Weekly claims data, published by the Department of Labor, are a separate indicator that analysts follow for timely signals of layoffs.

The broader measures

The official rate is only one measure. The BLS publishes a set of alternative measures of labor underutilization every month. They range from a narrow measure that counts only people unemployed for 15 weeks or more, to a broad measure that adds marginally attached workers, including discouraged workers, and people working part time who would prefer full-time work for economic reasons. The broadest of these is widely known as U-6, and the official rate is known as U-3. The BLS tables use these labels in the Employment Situation release, which is where to find the current values.

Because the broad measure includes people who want more work than they have, it is always higher than the official rate. The gap between the two can widen or narrow over the cycle, and economists use it to assess slack that the headline rate misses.

Household survey versus payroll survey

The monthly Employment Situation report combines two surveys. The household survey, the CPS, produces the unemployment rate and the household count of employment. The establishment survey, formally the Current Employment Statistics (CES) program, asks a large sample of businesses and government agencies how many people are on their payrolls. It produces nonfarm payrolls, the headline job gains figure, along with hours and earnings.

The two measure different things. The payroll survey counts jobs, so a person with two jobs may be counted twice, and it excludes the self-employed and agricultural workers. The household survey counts people, includes the self-employed and farm workers, and is the only source for the unemployment rate. The household survey also has a smaller sample relative to the size of the population, which means monthly changes can be noisy. Both surveys are subject to sampling error, and both are periodically revised or re-benchmarked. The BLS publishes information on the statistical significance of changes in the release.

For a reader, the practical rule is to use payrolls for the direction of job growth and the household survey for the unemployment rate, participation and demographic detail, while recognizing that the two may diverge in any given month.

A worked example

Consider a simplified economy with 1,000 adults. Of these, 600 are employed, 40 are unemployed and 360 are not in the labor force. The labor force is 640. The unemployment rate is 40 divided by 640, or 6.25 percent. The participation rate is 640 divided by 1,000, or 64 percent.

Now suppose 20 of the unemployed give up looking and move out of the labor force, with no one finding work. The labor force falls to 620, unemployed to 20, and the unemployment rate drops to about 3.2 percent. Employment has not changed at all. Participation dropped to 62 percent. This is a stylized example, but it shows why a falling rate is not automatically good news about hiring.

Reading the monthly report

The BLS releases the Employment Situation report monthly, and its schedule is posted on bls.gov. The report starts with a summary of the unemployment rate and nonfarm payroll change, followed by tables on demographic groups, duration of unemployment, reasons for unemployment and part-time work. A few points help:

• Check participation and the employment-population ratio, not just the rate.
• Look at the reasons for unemployment: job losers, job leavers, reentrants and new entrants point to different dynamics.
• Compare the household and payroll signals, and consider the revision to prior months in the payroll data.
• Remember that small monthly changes in the rate may be within the margin of sampling error.

FAQ

Does the unemployment rate count people who gave up looking?

No. People who are not actively searching are classified as outside the labor force. Discouraged workers are counted in the broader underutilization measures but not in the official rate.

Who conducts the survey?

The Census Bureau conducts the CPS for the BLS. Interviewers contact roughly 60,000 households a month.

Is the rate based on unemployment benefit claims?

No. It is based on survey responses. Benefit claims are a separate data series published by the Department of Labor.

What is U-6?

U-6 is the BLS’s broadest alternative measure of labor underutilization, adding marginally attached workers and people working part time for economic reasons to the unemployed.

Why do payrolls and the unemployment rate sometimes disagree?

They come from different surveys that count different things: jobs reported by employers versus people reported by households.

What counts as employed?

Anyone who did any paid work in the reference week, or 15 hours or more of unpaid work in a family business, plus people temporarily away from a job.