How the Government Measures Unemployment

U.S. Department of Labor
Robert B. Reich, Secretary

Bureau of Labor Statistics
Katharine G. Abraham, Commissioner
February 1994

Report 864

The Bureau of Labor Statistics receives many inquiries regarding the methods used to measure unemployment. This publication answers some of the most frequently raised questions about who is counted as unemployed and how unemployment is measured. This report was prepared in the Office of Employment and Unemployment Statistics by John F. Stinson, Jr. under the direction of Gloria Peterson Green, Chief, Data Users' and Publication Services Group. It is an updated version of the publication of the same title originally issued in 1964 and revised several times subsequently. For technical information on any aspects of unemployment, call (202) 606-6378, or write to the Division of Labor Force Statistics, 2 Massachusetts Ave., N.E., Room 4675, Washington, D.C. 20212-0001. Material in this publication is in the public domain and, with appropriate credit, may be reproduced without permission. This information will be made available to sensory impaired individuals upon request. Voice phone: (202) 606-STAT, TDD phone: (202) 606-5897, TDD message referral phone: 1-800-326-2577.

Chart 1. Civilian employment and unemployment, January-December 1993..
Table 1. Employment status of the civilian noninstitutional
population, 1993 annual averages......................................

Why does the Government collect statistics on the unemployed?

When workers are unemployed, they, their families and the country as a whole lose. Workers and their families lose wages, and the country loses the goods or services which could have been produced. In addition, the purchasing power of these workers is lost, which can lead to unemployment for yet other workers. To know about unemployment—the extent and nature of the problem—requires information. How many people are unemployed? How did they become unemployed? How long have they been unemployed? Are their numbers growing or declining? Are they men or women? Are they young or old? Are they white or black or of Hispanic origin? Are they skilled or unskilled? Are they the sole support of their families, or do other family members have jobs? Are they more concentrated in one area of the country than another? After these statistics are obtained, they have to be interpreted properly so they can be used—together with other economic data—by policymakers in making decisions as to whether measures should be taken to influence the future course of the economy or to aid those affected by joblessness.

Where do the statistics come from?

Early each month, the Bureau of Labor Statistics (BLS) of the U.S. Department of Labor announces the total number of employed and unemployed persons in the United States for the previous month, along with many characteristics of such persons. These figures, particularly the unemployment rate—which tells you the percent of the labor force that is unemployed—receive wide coverage in the press, on radio, and on television.

Some people think that to get these figures on unemployment the Government uses the number of persons filing claims for unemployment insurance (UI) benefits under State or Federal Government programs. But some people are still jobless when their benefits run out, and many more are not eligible at all or delay or never apply for benefits. Indeed, typically less than a third of the unemployed file claims for UI benefits. So, quite clearly, UI information cannot be used as a source for complete information on the number of unemployed.

Other people think that the Government counts every unemployed person each month. To do this, every home in the country would have to be contacted—just as in the population census every 10 years. This procedure would cost way too much and take far too long. Besides, people would soon grow tired of having a census taker come to their homes every month, year after year, to ask about job-related activities.

Because unemployment insurance records relate only to persons who have applied for such benefits, and since it is impractical to actually count every unemployed person each month, the Government conducts a monthly sample survey called the Current Population Survey (CPS) to measure the extent of unemployment in the country. The CPS has been conducted in the United States every month since 1940 when it began as a Work Projects Administration project. It has been expanded and modified several times since then. As explained later, the CPS estimates, beginning in 1994, reflect the results of a major redesign of the survey.

There are about 60,000 households in the sample for this survey. The sample is selected so as to be representative of the entire population of the United States. In order to select the sample, first, the 3,137 counties and county-equivalent cities in the country are grouped into 1,973 geographic areas. The Bureau of the Census then designs and selects a sample consisting of 729 of these geographic areas to represent each State and the District of Columbia. The sample is a State-based design and reflects urban and rural areas, different types of industrial and farming areas, and the major geographic divisions of each State.

Each of the 729 areas in the sample is subdivided into enumeration districts of about 300 households. The enumeration districts, in turn, are divided into smaller clusters of about four dwelling units each, through the use of address lists, detailed maps, and other sources. Then, the clusters to be surveyed are chosen statistically, and the households in these clusters are interviewed. The approximately 60,000 households eligible for interview each month represent about 1 in every 1,600 households in the country.

Each month, one-fourth of the households in the sample are changed, so that no household is interviewed more than 4 consecutive months. This practice avoids placing too heavy a burden on the households selected for the sample. After a household is interviewed for 4 consecutive months, it leaves the sample for 8 months and then is again interviewed for the same 4 calendar months a year later, before leaving the sample for good. This procedure results in approximately 75 percent of the sample remaining the same from month to month and 50 percent from year to year.

Each month, 1,500 highly trained and experienced Census Bureau employees interview persons in the 60,000 sample households for information on the labor force activities (jobholding and jobseeking) or non-labor force status of the members of these households during the week that includes the 12th of the month (the reference week). This information, relating to all household members 16 years of age and over, is entered by the interviewers into laptop computers; at the end of each day's interviewing, the data collected are transmitted to the Census Bureau's central computer in Washington, D.C. In addition, a portion of the sample is interviewed by phone through two central data collection facilities. (Prior to 1994, the interviews were conducted using a paper questionnaire which had to be mailed in by the interviewers each month.)

Each person is classified according to the activities he/she engaged in during the reference week. Then, the total numbers are "weighted," or adjusted to independent population estimates (based on updated decennial census results). The weighting takes into account the age, sex, race, Hispanic origin, and State of residence of the population, so that these characteristics are reflected in the proper proportions in the final estimates.

A sample is not a total count and the survey may not produce the same results that would be obtained from interviewing the entire population. But the chances are 90 out of 100 that the monthly estimate of unemployment from the sample is within 230,000 of the figure obtainable from a total census. Since monthly unemployment totals have ranged between about 6 and 11 million in recent years, the possible error resulting from sampling is not large enough to distort the total unemployment picture.

Because these interviews are the basic source of data for total unemployment, information must be factual and correct. Respondents are never asked specifically if they are unemployed, nor are they given an opportunity to decide their own labor force status. Unless they already know how the Government defines unemployment, many of them may not be sure of their actual classification when the interview is completed.

Similarly, interviewers do not decide the respondents' labor force classification. They simply ask the questions in the prescribed way and record the answers. Individuals are then classified as employed or unemployed by the computer based on the information collected and the definitions programmed into the computer.

All interviews must follow the same procedures to obtain comparable results. Because of the crucial role interviewers have in the household survey, a great amount of time and effort is spent maintaining the quality of their work. Interviewers are given intensive training, including classroom lectures, discussion, practice, observation, home-study materials, and on-the-job training. At least once a year, they convene for day-long training and review sessions, and, also at least once a year, they are accompanied by a supervisor during a full day of interviewing to determine how well they carry out their assignments.

A selected number of households are reinterviewed each month to determine whether the information obtained in the first interview was correct. The information gained from these reinterviews is used to improve the entire training program.

What are the basic concepts of employment and unemployment?

The basic concepts involved in identifying the employed and unemployed are quite simple:

  • People with jobs are employed.
  • People who are jobless, looking for jobs, and available for work are unemployed.
  • People who are neither employed nor unemployed are not in the labor force.

The survey is designed so that each person age 16 and over who is not in an institution such as a prison or mental hospital or on active duty in the Armed Forces is counted and classified in only one group. The sum of the employed and the unemployed constitutes the civilian labor force. Persons not in the labor force combined with those in the civilian labor force constitute the civilian noninstitutional population 16 years of age and over. Under these concepts, most people are quite easily classified. For example:

Elizabeth Lloyd reported to the interviewer that last week she worked 40 hours as a sales manager for the Western Beverage Company.

Steve Hogan lost his job when the local plant of the Chariot Aircraft Manufacturing Company was closed down. Since then, he has been visiting the personnel offices of the other factories in the town trying to find a job.

Linda Coleman is a homemaker. Last week, she was occupied with her normal household chores. She neither held a job nor looked for a job. Her 80-year old father who lives with her has not worked or looked for work because of a disability.

Each of these examples is clear cut. Elizabeth is employed; Steve is unemployed; and Linda and her father are not in the labor force.

Who is counted as employed?

Not all of the wide range of job situations in the American economy fit neatly into a given category. For example, people are considered employed if they did any work at all for pay or profit during the survey week. This includes all part-time and temporary work, as well as regular full-time year-round employment. Persons also are counted as employed if they have a job at which they did not work during the survey week because they were:

  • On vacation;
  • Ill;
  • Experiencing child-care problems;
  • Taking care of some other family or personal obligation;
  • On maternity or paternity leave;
  • Involved in an industrial dispute; or
  • Prevented from working by bad weather.

These persons are counted among the employed and tabulated separately as "with a job but not at work," because they have a specific job to which they will return.

But what about the two following cases? George Lewis is 16 years old, and he has no job from which he receives any pay or profit. However, George does help with the regular chores around his father's farm about 20 hours each week.

Lisa Fox spends most of her time taking care of her home and children, but, all day Friday and Saturday, she helps in her husband's computer software store.

Under the Government's definition of employment, both George and Lisa are considered employed. They fall into a group called "unpaid family workers," which includes any person who worked 15 hours or more in a week without pay in a family-operated enterprise. Such persons contribute significantly to our productive effort and are an important part of our labor supply, particularly in agriculture and retail trade. However, unpaid family workers who work fewer than 15 hours per week are counted as "not in the labor force."

Who is counted as unemployed?

Persons are classified as unemployed if they do not have a job, have actively looked for work in the prior 4 weeks, and are currently available for work. Actively looking for work may consist of any of the following activities:

  • Contacting:
    An employer directly or having a job interview;
    A public or private employment agency;
    Friends or relatives;
    A school or university employment center;
  • Sending out resumes or filling out applications;
  • Placing or answering advertisements;
  • Checking union or professional registers; or
  • Some other means of active job search.

Passive methods of jobsearch do not result in jobseekers actually contacting potential employers, and therefore are not acceptable for classifying persons as unemployed. These would include such things as attending a job training program or course or merely reading the want ads.

Workers expecting to be recalled from layoff are counted as unemployed, whether or not they have engaged in a specific jobseeking activity. But, in all other cases, the individual must be actively engaged in some job search activity and available for work (except for temporary illness).

The questions used in the interviews are carefully designed to elicit the most accurate picture of each person's labor force activities. Some of the major questions that determine employment status are: (The capitalized words are emphasized when read by the interviewers.) 1. Does anyone in this household have a business or farm? 2. LAST WEEK, did you do ANY work for (either) pay (or profit)? If the answer to question 1 is "yes" and the answer to question 2 is "no," the next question is: 3. LAST WEEK, did you do any unpaid work in the family business or farm? For those who reply "no" to both questions 2 and 3, the next key questions used to determine employment status are: 4. LAST WEEK, (in addition to the business,) did you have a job either full or part time? Include any job from which you were temporarily absent. and 5. LAST WEEK, were you on layoff from a job? and 6. What was the main reason you were absent from work LAST WEEK? For those who respond "yes" to question 5 about being on layoff, the following questions are asked: 7. Has your employer given you a date to return to work? and, if "no," 8. Have you been given any indication that you will be recalled to work within the next 6 months? If the responses to either question 7 or 8 indicate that the person expects to be recalled from layoff, he/she is counted as unemployed. For those who were reported as having no job or business from which they were absent or on layoff, the next question is: 9. Have you been doing anything to find work during the last 4 weeks? For those who say "yes," the next question is: 10. What are all of the things you have done to find work during the last 4 weeks? If an active method of looking for work, such as those listed at the beginning of this section, is mentioned, the following question is asked: 11. LAST WEEK, could you have started a job if one had been offered? If there is no reason, except temporary illness, that the person could not take a job, he/she is considered to be not only looking but also available for work and is counted as unemployed.

Some examples of responses that are typically given in interviews and that may result in a person being classified as unemployed are: 1. Yvonne Bennett reported that 2 weeks ago she applied for a job as a receptionist at the Capitol Travel Agency and the Equity Mortgage Lending Company. She is awaiting the results of her applications. Yvonne is unemployed because she made a specific effort to find a job within the prior 4 weeks and is presently available for work. 2. Mrs. Jenkins tells the interviewer that her daughter Katherine Marie was thinking about looking for work in the prior 4 weeks but knows of no specific efforts she has made. Katherine Marie does not meet the activity test for unemployment and is, therefore, counted as "not in the labor force." 3. Paul Flynn has been checking for openings on a bricklayer's union register for each of the past 3 weeks, but his wife reported that last week he had the flu and was unable to work because of it. Paul is counted as unemployed because he took steps to look for work in his trade and would have been available for work during the survey reference week, except for his temporary illness. 4. Marcus Green was laid off from the Hotshot Motor Company when the firm began retooling to produce a new model car. Marcus knows he will be called back to work as soon as the model changeover is completed, and he also knows it is unlikely that he would be able to find a job for the period he is laid off; so, although he is available to work, he is not seeking a job. Marcus is unemployed because he is waiting to be recalled from layoff. 5. Joan Howard told the interviewer that she has filed applications with three companies for summer jobs. However, it is only April and she doesn't wish to start work until at least June 15, because she is attending school. Although she has taken specific steps to find a job, Joan is classified as not in the labor force because she is not currently available for work. From these definitions and examples, it can be seen that the total unemployment figures cover more than the number of persons who have lost jobs. It includes persons who have quit their jobs to look for other employment, workers whose temporary jobs have ended, persons looking for their first jobs, and experienced workers looking for jobs after an absence from the labor force (as, for example, a woman who returns to the labor force after her children have entered school).

Sometimes, questions are raised as to the advisability of including among the unemployed persons such as teenagers, particularly those who are going to school, and others who are not usually the family's main supporter. The point is made that, for the family, unemployment of the person maintaining the household is more important than the unemployment of any other member of the family.

This view ignores some fundamental changes that have occurred both in life-styles and attitudes toward work. For example, the "traditional family" consisting of a working husband, a wife who is not in the labor force, and one or more children accounts for only about 1 out of 10 American families. In about one-fifth of married-couple families, the husband is no longer an earner. Nearly one-fifth of the unemployed wives in 1993 were in families where no one was working.

In 1993, 30 percent of the labor force consisted of married women and teenagers. They are counted as employed when they contribute to the total productive effort of the Nation. Otherwise, there would be no way to measure the total amount of labor resources the economy is using. If persons are counted as employed when they have a job, no matter how many hours they might be working, they must be counted as unemployed when they have no job and are searching for one; otherwise, the measurement of unemployment loses much of its meaning, and a very important indication of business conditions is lost. Unemployment statistics are intended to provide counts of unused, available labor resources. They are not measures of the number of persons who are suffering economic hardship.

Who is not in the labor force?

Persons under 16 years of age are automatically excluded from the official labor force measurements, as are all inmates of institutions and persons on active duty in the Armed forces. All other members of the civilian noninstitutional population are eligible for inclusion in the labor force, and those 16 and over who have a job or are actively looking for one are so classified. All others—those who have no job and are not looking for one—are counted as "not in the labor force." Many who do not participate in the labor force are going to school or are retired. Family responsibilities keep others out of the labor force. Still others have a physical or mental disability which prevents them from participating in labor force activities.

A series of questions is asked each month of persons not in the labor force to obtain information about their desire for work, the reasons why they had not looked for work in the last 4 weeks, their prior job search, and their availability for work. These questions include: 1. Do you currently want a job, either full or part time? 2. What is the main reason you were not looking for work during the LAST 4 WEEKS? 3. Did you look for work at any time during the last 12 months? 4. LAST WEEK, could you have started a job if one had been offered?

These questions form the basis for estimating the number of "discouraged workers"—persons without jobs who are not currently looking for work (and, therefore, are not counted among the unemployed) but who, nevertheless, have demonstrated some degree of labor force attachment. Specifically, to be classified as discouraged workers, individuals must indicate that they currently want a job, have looked for work in the last 12 months (or since they last worked, if they had worked sometime in the last year), are available for work, and that they are not currently looking for work either because they believe that no work is available in their line of work or area, they had previously been unable to find work, they lack necessary schooling, training, skills or experience, employers think they are too young or old, or they face some other type of discrimination.

Additional questions about persons not in the labor force are asked of each household during the last month of its 4-month tenure in the sample rotation pattern. These questions are designed to collect information about why these people left their previous jobs, when they last worked at a job or business, whether they intend to look for work in the near future, and their current situation (retired, disabled, ill, in school, taking care of house or family, or something else).

What about cases of overlap?

When the population is classified according to who is employed, unemployed, and not in the labor force on the basis of their activities during a given calendar week, situations are often encountered where individuals have engaged in more than one activity. Since persons are counted only once, it must be decided which activity will determine their status. Therefore, a system of priorities is used: Labor force activities take precedence over non-labor force activities. Working or having a job takes precedence over looking for work. Some examples are: 1. James Kelly and Elyse Martin attend Jefferson High School. James works after school at the North Star Cafe, and Elyse is seeking a part-time job at the same establishment (also after school.) James' job takes precedence over his non-labor force activity of going to school, as does Elyse's search for work; therefore, James is counted as employed and Elyse is counted as unemployed. 2. Last week, Mary Davis, who was working for Stuart Comics, went to the Coastal Video Shop on her lunch hour to be interviewed for a higher paying job. Mary's interview constitutes looking for work, but her work takes priority, and she is counted as employed. (Indeed, because of the way the questionnaire is set up, information about Mary's search for work is not even obtained.) 3. John Hinton has a job at the Nuts and Bolts Company, but he didn't go to work last week because of a strike at the plant. Last Thursday, he went to the Screw and Washer Factory to see about a temporary job until the strike terminates. John was "with a job but not at work" due to an industrial dispute, which takes priority over looking for work; therefore, he is counted as employed. (Again, information would not be obtained on John's jobsearch effort.) 4. Amy Douglas lost her job at the book store on Wednesday of the survey week. She answered newspaper want-ads on Thursday and Friday but had not obtained a new job by the end of the week. Amy is counted as employed, since she was paid for the 3 days that she did work, even though she was unemployed for part of the survey week. (Once again, information would not be obtained on her search for work, though Amy would be identified as working "part time for economic reasons," by virtue of her having her workweek reduced to part time by her dismissal from her previous job.)

To summarize: Employed persons consist of:

  • All persons who did any work for pay or profit during the survey week.
  • All persons who did at least 15 hours of unpaid work in a family-operated enterprise.
  • All persons who were temporarily absent from their regular jobs because of illness, vacation, bad weather, industrial dispute, or various personal reasons.

    Unemployed persons are:

  • All persons who did not have a job at all during the survey week, made specific active efforts to find a job during the prior 4 weeks, and were available for work (unless temporarily ill).
  • All persons who were not working and were waiting to be called back to a job from which they had been temporarily laid off.

    Because of the complexities of the American economic system and the wide variety of job arrangements and jobseeking efforts, the definitions of employment and unemployment must be specific so as to ensure uniformity of reporting at any given time and over any period of time. When all of the details are considered, definitions may seem rather complicated. The basic concepts, however, are still the same: People with jobs are employed, people who do not have jobs and are looking for jobs are unemployed, and people who meet neither labor market test are not in the labor force. The qualifying conditions are necessary to cover the wide range of labor force patterns and to provide an objective set of standards for consistent treatment of cases.

    How large is the labor force?

    Now that the major labor force concepts have been defined, it is useful to see how they apply to the civilian noninstitutional population in a specific year.

    Table 1 indicates that the majority of civilians age 16 years and over are in the labor force. Of the 193.6 million persons in the civilian noninstitutional population in 1993, 128.0 million were either working or looking for work, that is, they were in the civilian labor force. The vast majority of those in the civilian labor force were employed, and most of the employed usually worked full time (35 hours or more).

    Table 1.  Employment status of the civilian noninstitutional population, 1993 annual averages

    (Thousands of persons 16 years of age and over)
    Employment Status | 1993 |

    Civilian noninstitutional population.........| 193,550 |

    Civilian labor force.......................| 128,040 Employed..................................| 119,306 At work..................................| 113,278 Full time.............................| 91,868 Part time.............................| 21,410 With a job but not at work...............| 6,028 Industrial dispute....................| 24 On vacation...........................| 3,330 Bad weather...........................| 151 Temporary illness.....................| 1,290 Other reasons.........................| 1,233 Unemployed................................| 8,734 Looking for full-time work............| 7,146 Looking for part-time work............| 1,588 |

    Not in the labor force.....................| 65,509

    Chart 1 shows trends in total employment and unemployment over the course of 1993. As shown, total employment and unemployment are normally higher in some parts of the year than in others. For example, unemployment is higher in January and February, when it is very cold in many parts of the country and work in agriculture and construction and other seasonal industries is curtailed. Also, both employment and unemployment rise every June, when students enter the labor force in search of summer jobs.

    The labor force, then, is not a fixed number of people. It increases with the long-term growth of the population, it responds to economic forces and social trends, and its size changes with the seasons.

    How are seasonal fluctuations taken into account?

    As suggested in the previous section, the number of employed and unemployed persons fluctuates during the year in a pattern that tends to repeat itself year after year and which reflects holidays, vacations, harvest time, seasonal shifts in industry production schedules, and similar occurrences. Because of such patterns, it is often difficult to tell whether developments between any 2 months reflect changing economic conditions or merely normal seasonal fluctuations. To deal with such problems, a statistical technique called seasonal adjustment is used.1

    When a statistical series has been seasonally adjusted, the normal seasonal fluctuations are smoothed out and data for any month can be more meaningfully compared with data from any other month or with an annual average. One of the most important seasonally adjusted figures that is published monthly is the unemployment rate, which represents total seasonally adjusted unemployment as a percent of the seasonally adjusted civilian labor force. Typically, the unadjusted rate peaks in January or February and reaches a low point in May or October. However, when the rate is seasonally adjusted and the underlying employment situation has not changed, the rate will be essentially the same throughout the year, varying only within a narrow range that results from sampling and other measurement errors. On the other hand, if changes occur in the underlying employment situation in addition to seasonal movements, these changes will show up as a rising or declining trend in the seasonally adjusted unemployment rate.

    What do the unemployment insurance figures measure?

    Statistics on insured unemployment in the United States are collected as a byproduct of unemployment insurance (UI) programs. Workers who lose their jobs and are covered by these programs typically file claims which serve as notice that they are beginning a period of unemployment. Claimants who qualify for benefits are counted in the insured unemployment figures.

    Some countries base their estimates of total unemployment on the number of persons filing claims for or receiving UI payments or the number of persons registered with government employment offices as available for work. These data are also available in the United States, but they are not used to measure total unemployment because they exclude several important groups. In terms of employed workers, the principal groups not covered are self-employed workers, unpaid family workers, workers in certain not-for-profit organizations, and several other, primarily seasonal, worker categories.

    In addition to those unemployed workers whose last jobs were in the excluded kinds of employment, the insured unemployed exclude the following: 1. Unemployed workers who have exhausted their benefits; 2. Unemployed workers who have not yet earned benefit rights (new entrants or reentrants to the labor force); 3. Disqualified workers whose unemployment is considered to have resulted from their own actions rather than from economic conditions, for example, a worker discharged for misconduct on the job; and 4. Otherwise eligible unemployed persons who do not file for benefits. Because of these and other limitations, statistics on insured unemployment, although having many important uses (one of which is discussed below), cannot be used as a count of total unemployment in the United States. In 1988, for example, when there were virtually no extended unemployment benefits paid to persons who had otherwise exhausted their benefits, the number receiving UI benefits represented only 31 percent of the total unemployed. In 1992, when extended UI benefits were in effect, this proportion was 51 percent.

    How is unemployment measured for States and local areas?

    While the Current Population Survey produces reliable monthly estimates of unemployment for the country as a whole, and annual estimates for all States, the District of Columbia, and selected metropolitan areas, its sample size is not large enough to produce reliable monthly estimates for most States and areas. Currently, monthly unemployment estimates are obtained directly from the CPS for 11 large States (California, Florida, Illinois, Massachusetts, Michigan, New Jersey, New York, North Carolina, Ohio, Pennsylvania, and Texas) and two areas, the Los Angeles-Long Beach metropolitan area and New York City.

    Monthly unemployment estimates for the remaining 39 States, the District of Columbia, and some 5,600 geographic areas, which include all labor market areas (LMA's)2 and counties and cities having a population of 25,000 or more, are developed by State employment security agencies using concepts, definitions, and estimating procedures established by BLS.

    For the 39 smaller States and the District of Columbia, which do not use the CPS directly each month, statistical techniques (regression models) are used to develop employment and unemployment estimates. The regression techniques used in these models are based on historical and current relationships within each State's economy as reflected in the different sources of data that are available, such as the CPS, the BLS survey of employers, and the State unemployment insurance (UI) system. While all of the State models have the same key explanatory variables, the coefficients for these variables are unique to each State, so as to better reflect individual State characteristics.

    The employment models use the historical relationship between the State's monthly employment estimates from the BLS survey of employers and the CPS. The models also include trend and seasonal components to account for movements in the CPS not captured by the business survey. Similarly, the unemployment rate models use the historical relationship between the State's monthly UI claims and its unemployment as measured by the CPS, along with trend and seasonal components.

    Once each year, usually with the release of January data, the monthly estimates generated by the models are adjusted, or benchmarked, by BLS to the annual average CPS estimates. The benchmarking employs a procedure which adjusts the annual average of the models to equal the CPS annual average, while largely preserving the original monthly seasonal pattern of the model estimates. No benchmark correction is required for the 11 States for which estimates are obtained directly from the CPS.

    Monthly employment and unemployment estimates for all sub-State areas except New York City and the Los Angeles-Long Beach metropolitan area are not created from models. Instead, they are derived through a building block procedure which categorically accounts for employment and unemployment in a given LMA. In this building block procedure, the two categories of employment are: (1) Persons employed in industries covered by State UI laws and (2) persons employed in industries not covered by those laws. The unemployment estimate consists of three groups of unemployed persons, two of which parallel the employment categories. They are: (1) Persons previously employed in industries covered by State UI laws, (2) persons previously employed in industries not covered by these laws, and (3) persons entering the civilian labor force for the first time or reentering after a period of time spent out of the labor force.

    These categories are aggregated to obtain estimates of employment and unemployment in the LMA. Once the current month's preliminary estimates are obtained, a proportional adjustment is applied to all LMA estimates to ensure that they sum to the State total. Estimates for counties, cities, and other sub-LMA areas are disaggregated from LMA estimates using UI claims, current total population, and decennial census data.

    At the end of each year, the sub-State unemployment estimates are revised. The revisions incorporate any changes in the inputs, such as corrections in UI claims counts and updated historical relationships. The corrected estimates are then readjusted to add to the revised (benchmarked) State unemployment estimates.3

    Is there a measure of underemployment?

    The fruits of the potential labor of the unemployed are lost to them and to the economy. Similarly, for people who are unable to work as many hours each week as they would like, their potential output and earnings for the hours in which they were willing but unable to work are also lost. Such workers are sometimes considered "underemployed."

    Because of the difficulty of developing an objective set of criteria which could be readily used in a monthly household survey, no official government statistics are available on the total number of persons who might be viewed as underemployed. Even if many or most could be identified, it would still be difficult to quantify the loss to the economy of such underemployment.

    The Government does collect data on workers who wish to work full time but, because of business and economic conditions, worked fewer than 35 hours during the survey week. There are two main types: The first consists of persons who usually work full time but whose hours had been temporarily cut back because of slack work or unfavorable business conditions. The second consists of persons who usually work fewer than 35 hours because they had been unable to find full-time work or because of slack work or business conditions.

    Underemployment in the form of shortened hours is relatively easy to measure. Other kinds of underemployment, though easy to illustrate, are not so easily quantified. One kind of underemployment occurs when persons take jobs that do not make use of, or pay according to, their skills, training, and experience. An extreme example would be an aerospace engineer working in a fast-food restaurant because no engineering position was available. Another kind of underemployment occurs when an individual does not have the necessary equipment to achieve maximum efficiency and output. A farmer who uses a mule and a handplow may work longer and harder and yet may harvest a much smaller crop than a neighbor who works less but has a modern tractor.

    Underemployment in a highly industrialized country like the United States is not the serious problem that it is for most developing countries. Nevertheless, to the extent that it does occur, underemployment represents a waste of resources and, therefore, is a subject of continuing interest.

    BLS has published a series of reports since 1982 that look beyond unemployment and examine other labor market problems faced by workers and assess the impact of these problems on their economic status and that of their families and households. In addition to unemployment, the reports focus on the issues of involuntary part-time employment and low earnings among full-time workers in the labor force for more than half of the year. Statistics for persons with these problems are then linked with data on income to determine how many of these people have incomes below the poverty level and/or live in families with incomes below the poverty level.4

    Have there been any changes in the definition of unemployment?

    The concepts and definitions underlying the labor force data have been modified, but not substantially altered, even though they have been under almost continuous review by interagency governmental groups, congressional committees, and private groups since the inception of the Current Population Survey.5

    In January 1994, a major redesign of the Current Population Survey was introduced which included a complete revamping of the questionnaire, the use of computer-assisted interviewing for the entire survey, and revisions to some of the labor force concepts.6

    The CPS was redesigned primarily to improve the quality of labor market information. The former CPS questionnaire had changed little since the last major revision in 1967. During the 1970's and 1980's, problems in measuring certain labor market concepts were identified. Also, U.S. society and the economy had undergone major changes during this period, such as the growth of service jobs and the decline of factory jobs; the more prominent role of women, especially mothers, in the work force; and a profusion of alternative work schedules. These changes raised new issues about labor force behavior which were being inadequately addressed by the former questionnaire. Finally, there have been major advances in survey research methods and data collection technology. Since the need for relevant, precise, and comprehensive data about the labor force continued to grow, the time was right to implement a state-of-the-art, fully computerized CPS.

    The redesign of the CPS was guided by the results of research projects conducted by BLS and the Bureau of the Census beginning in 1988 as well as the recommendations made in 1979 by the National Commission on Employment and Unemployment Statistics (NCEUS) in their report, Counting the Labor Force. Some of the major changes to the survey were:

    1. The introduction of a redesigned and automated questionnaire. The questionnaire was totally revamped in order to obtain more accurate, comprehensive, and relevant information, to implement several definitional changes, and to institute a completely automated interview. Instead of using paper questionnaires, CPS interviewers are now equipped with laptop computers which are programmed with the questionnaire.7 The automation of the questionnaire has many benefits, such as allowing for complex skip patterns among the questions; built-in followup questions; customized questions; and dependent interviewing, that is, the use in the current interview of information obtained during a previous month's interview. The use of the new questionnaire also reduces the possibility of the misclassification of a person's labor force status by sharpening the measurement of the labor force concepts and making the questions easier to understand and answer.

    2. The addition of two more restrictive criteria to the definition of discouraged workers, whereby, to be counted as discouraged, persons must have looked for a job within the past year (or since their last job, if they worked during the year) and must indicate that they were available to start a job during the survey reference week if one had been offered. As in the past, the discouraged still have to state that they want a job now and that they did not look for work in the prior 4 weeks for one of the following reasons: Believes no work available in line of work or area; had previously not been able to find any work; lacks necessary schooling, training, skills, or experience; employers think too young or old; or other types of discrimination. The previous definition of discouraged workers was thought by the NCEUS and others to be too subjective, because it was based primarily on an individual's stated desire for a job. Additionally, the discouraged worker data are now collected from the full CPS sample and published on a monthly basis; prior to 1994, these data were collected from only a quarter of the CPS sample and published quarterly.

    3. The addition of questions about the expectation of recall for persons who indicate that they are on layoff. A person on layoff is defined as someone who was separated from a job and expects to be recalled. In the past, the questionnaire did not include direct questions concerning persons' expectations of recall.

    4. A new requirement that persons waiting to start a new job within 30 days meet the jobsearch requirement of having actively looked for work in the prior 4 weeks in order to be classified as unemployed. Otherwise, they are classified as not in the labor force. Previously, they did not have to meet the jobsearch requirement in order to be included among the unemployed. This change effectively eliminates this numerically small category as a separately identifiable group, as they are no longer an exception to the jobsearch requirement.

    5. Sharpening the measurement of unemployment by making subtle but important changes to the questions on jobsearch. To help the CPS interviewers distinguish between active jobsearch methods (which qualify a person as unemployed) and passive methods (which do not qualify), the response categories were expanded and reformatted.

    6. The introduction of changes to the question on the reasons for unemployment to include a response category for people who became unemployed when their temporary job ended. Also, the criteria for determining whether an unemployed person should be classified as a new or reentrant to the labor force have been simplified; now, any prior work experience, including part-time work, qualifies an unemployed person to be a "reentrant."

    7. Allowing respondents to report the duration of their jobsearch in either weeks, months, or years, whichever they prefer. Previously, they had to answer in weeks. In a form of dependent interviewing, the information on the duration of jobsearch will be collected in the first month the person is unemployed and then will be automatically updated in the subsequent months, as long as the person remains unemployed.

    8. Tightening the measurement of persons employed part time for economic reasons (those who would prefer full-time work but are limited, at least temporarily, to part-time schedules) by adding two criteria for persons who usually work part time: They must be explicitly reported as wanting a full-time job and as having been available to take one during the reference week. (Persons who usually work full time but worked part time for an economic reason during the reference week are assumed to meet these criteria.) Previously, such information was inferred.

    The most relevant results of these changes were: An increase in the number of persons reporting some labor force activity, particularly among women; an increase in the unemployment rate, again largely as a reflection of increased reporting of job-search activities by women; a sharp reduction in the number of persons classified as discouraged workers; and a reduction in the number of workers classified as working part time for economic reasons.8

    Although it is impossible to determine the exact extent to which these series and other labor force estimates were affected by the January 1994 changes, BLS is doing several things to help users interpret the data from the new survey. First, information from a parallel survey of 12,000 households using the new survey questionnaire and methodology which was conducted from July 1992 to December 1993 has been used to develop estimates of what the unemployment rate would have been in 1993 had the new survey been used. Secondly, model-based estimates of the unemployment rate have been developed to estimate what the rate would have been had the old survey been continued.9 Also, a follow-on sample survey of 12,000 households that uses the old questionnaire and methods was introduced in January 1994; data from this survey will become available for additional comparisons in mid-1994. Finally, research is being conducted which will lead to the production of a variety of historical series for use by analysts and researchers.

    How are the unemployed counted in other countries?

    The sample survey system of counting the unemployed in the United States is also used by many foreign countries, including Canada, Mexico, Australia, Japan, and all of the countries in the European Economic Community. More recently, a number of Eastern European nations have instituted labor force surveys as well. However, some countries collect their official statistics on the unemployed from employment office registrations or unemployment insurance records. Many nations, including the United States, use both labor force survey data and administrative statistics to analyze unemployment.

    Sample surveying is the standard method recommended by international committees of specialists, under the auspices of the International Labour Office (ILO). In October 1982, the ILO's 13th International Conference of Labour Statisticians adopted revised international standards for statistics on the labor force, employment, and unemployment.10 The United States follows the international definitions, but the U.S. definitions are somewhat more specific.

    BLS regularly compiles a series of unemployment rates, with adjustments to U.S. concepts, for nine foreign nations—Australia, Canada, France, Germany, Italy, Japan, the Netherlands, Sweden, and the United Kingdom.11 International differences in jobless rates, even after adjustment to a common conceptual framework, only partially reflect differences in economic performance. Each country's combination of social and institutional programs also contributes to the differentials. Demographic factors such as growth and composition of the labor force also affect the unemployment rates.

    Who are the unemployed?

    The figure for total unemployment cannot possibly meet the requirements of all people who wish to use labor market data. Thus, on a monthly, quarterly, or annual basis the Bureau publishes in Employment and Earnings over 30,000 separate statistics showing employment and unemployment by age, sex, race, marital status, family relationship, industry, occupation, Vietnam-era veteran status; hours of work for the employed; and, for the unemployed, duration of unemployment, jobseeking methods, and the reasons for unemployment. Suppose one wants to know: How many of the unemployed are teenagers? How many of the unemployed are adult men and women? How many of the unemployed lost their last job? How many of the unemployed left their last job? What methods do the unemployed use to find work? How many of the unemployed have been seeking jobs for 27 weeks or more? How many of the unemployed are looking for their first job? How many of the unemployed are seeking part-time work? How many persons are too discouraged to look for a job? What industries have the highest rates of unemployment? What occupations have the highest rates of unemployment? How does the unemployment rate for black workers compare with the rates for white and Hispanic workers? How does the unemployment rate for Vietnam-era veterans compare with that for nonveterans? How many of the unemployed have no employed persons in their family? What States or areas have the highest rates of unemployment? How many of the employed are working part time when they want full-time jobs? The detailed figures which are published monthly, quarterly, and annually provide the answers to all of these questions and hundreds more. The depth and breadth of this information would be impossible to convey in totals alone—no matter how employment and unemployment are defined. These figures are as important as the totals in analyzing employment conditions and in making policy decisions. The total tells the number unemployed, but detailed statistics answer the question: Who are the unemployed?

    Selected Bibliography

    Bregger, John E. "The Current Population Survey: A Historical Perspective and BLS' Role," Monthly Labor Review, June 1984. Hamel, Harvey R. and Tucker, John T. "Implementing the Levitan Commission's Recommendations To Improve Labor Force Data," Monthly Labor Review, February 1985. McIntire, Robert J. "Revision of Seasonally Adjusted Labor Force Series," Employment and Earnings, January 1994. National Commission on Employment and Unemployment Statistics. Counting the Labor Force. Washington, D.C., 1979. President's Committee to Appraise Employment and Unemployment Statistics. Measuring Employment and Unemployment. Washington, D.C., 1962. U.S. Department of Labor, Bureau of Labor Statistics. Employment and Earnings. (Monthly publication containing current labor force data and explanatory notes on their underlying concepts, definitions, and collection methods.) U.S. Department of Labor, Bureau of Labor Statistics. Employment Situation news release. (Monthly news release in which the current labor force data are initially published.) 1 The official seasonal adjustment procedure for the labor force series is the X-11 ARIMA method, which was developed at Statistics Canada during the 1970's as an extension of and improvement to the widely used X-11 method developed at the U.S. Bureau of the Census in the 1960's. The primary documentation for the X-11 ARIMA procedure is The X-11 ARIMA Seasonal Adjustment Method, (Statistics Canada Catalogue No. 12-564E, January 1983). (ARIMA is an acronym for Auto-Regressive Integrated Moving Average.)

    2 A general definition for a labor market area is that it is an economically integrated geographic area within which individuals can reside and find employment within a reasonable distance or can readily change employment without changing their place of residence.

    3 A more detailed description of the methods used in developing unemployment estimates for States and areas appears in the Explanatory Notes of Employment and Earnings.

    4 For the most recent report in the series, see A Profile of the Working Poor, 1991, Report 847, U.S. Department of Labor, Bureau of Labor Statistics, June 1993.

    5 In January 1967, several labor force concepts and definitions were sharpened based on recommendations made by the President's Committee to Appraise Employment and Unemployment Statistics in their report, Measuring Employment and Unemployment. A detailed discussion of the changes made and their effect on the various series is contained in "New Definitions for Employment and Unemployment," Employment and Earnings and Monthly Report on the Labor Force, February 1967.

    6 A detailed discussion of the January 1994 changes, the reasons for them, and their possible impact can be found in three articles, "Overhauling the Current Population Survey—Why Is It Necessary to Change; Redesigning the Questionnaire; and Evaluating Changes in the Estimates," Monthly Labor Review, September 1993, pp.3-33. For a shorter discussion and description of the changes, see Questions and Answers on the Redesign of the Current Population Survey, Report 866, U.S. Department of Labor, Bureau of Labor Statistics, February 1994.

    7 Some of the interviewing is carried out from centralized computer centers of the Bureau of the Census, involving interviewers calling respondents from computer workstations.

    8 For a detailed discussion of the impact of the redesign on labor force estimates, see the article, "Revisions in the Current Population Survey Effective January 1994," in the February 1994 issue of Employment and Earnings.

    9 For preliminary findings from this research, see two BLS Technical Reports: "What would the Unemployment Rate have been had the Redesigned Current Population Survey been in Place from September 1992 to December 1993?: A Measurement Error Analysis," by Stephen M. Miller and "Predicting the National Unemployment Rate that the 'Old' CPS would have Produced," by Richard Tiller.

    10 For the full text of the 1982 resolution, see "Thirteenth International Conference of Labour Statistics," Bulletin of Labour Statistics (Geneva, International Labour Office, 1983-3), pp. ix-xxvi.

    11 For an in-depth description and comparison of the methods used to estimate unemployment in the United States and foreign countries, see International Comparisons of Unemployment, BLS Bulletin 1979, August 1978, and its statistical and technical supplements.


    Last Modified Date: October 16, 2001