Chapter 1.
Labor Force Data Derived from the Current Population Survey

Recent Changes to the Survey

Sample expansion
Beginning with the release of July 2001 data, labor force estimates from the CPS reflect the expansion of the monthly CPS sample from about 50,000 to about 60,000 eligible households. This expansion was one part of the Census Bureau’s plan to meet the requirement of the State Children’s Health Insurance Program (SCHIP) legislation. The SCHIP legislation requires the Census Bureau to improve State estimates of the number of children who live in low-income families and lack health insurance. These estimates are obtained from the Annual Demographic Supplement to the CPS (better known as the March income supplement).

In September 2000, the Census Bureau began expanding the monthly CPS sample in 31 states and the District of Columbia. The additional 10,000 households were added to the sample over a 3-month period. BLS chose not to include the additional households in the official labor force estimates, however, until it had sufficient time to evaluate the estimates from the expanded sample.

Estimates at the national level (not seasonally adjusted) derived from the 50,000- and 60,000-household samples were virtually the same. In any given month, the 60,000-household sample estimates for the overall labor force participation rate and the employment-population ratio differed by no more than 0.1 percentage point from estimates produced from the 50,000-household sample. The overall unemployment rates were identical in both samples. (For a discussion of the effect of the sample expansion on State estimates, see the forthcoming update of chapter 4.)

The 1994 redesign
A major redesign of the CPS was implemented in January 1994. The primary objective was to improve the quality of the data derived from the survey by introducing a new questionnaire and modernized data collection methods. Prior to 1994, the survey questionnaire had been virtually unchanged since 1967, at which time changes had been introduced based on recommendations of the Gordon Committee (President’s Committee to Appraise Employment and Unemployment Statistics, 1962). Additional changes were proposed in the late 1970s based on the recommendations of the Levitan Commission (National Commission on Employment and Unemployment Statistics, 1979); these, in part, formed the basis for the 1994 redesign.

The redesign of the questionnaire had four main objectives: 1) To adopt a computer-assisted interviewing environment, 2) to measure the official labor force concepts more precisely, 3) to expand the amount of data available, and 4) to implement several definitional changes.

Computerization. The new questionnaire was designed for a computer-assisted interview, in which interviewers ask the survey questions as they appear automatically on the screen of their laptop computer, and then type the responses directly into the laptop. In most cases, interviewers conduct the survey either in person at the respondent’s home or by telephone from the interviewer’s home. This mode of data collection is known as computer-assisted personal interviewing (CAPI). (In addition, about 10 percent of sample households are interviewed from centralized telephone centers, as explained below in the section on collection methods.)

Computer-assisted interviewing has important benefits, most notably that it facilitates the use of a relatively complex questionnaire that incorporates complicated skip patterns and standardized followup questions. Additionally, certain questions are automatically tailored to the individual’s situation to make them more understandable. The computerized questionnaire also has several built-in editing features, including automatic checks for internal consistency and unlikely responses. An automated interview also permits dependent interviewing, that is, the use of information in the current interview that was obtained in a previous month’s interview. Dependent interviewing reduces respondent and interviewer burden, while improving consistency of the data from one month to the next. The technique is being used to confirm the previously reported industry and occupation of a person’s job, to calculate unemployment duration, and, for many people not in the labor force, to confirm their status as retired or disabled.

Major questionnaire changes. While the labor force status of most people is straightforward, some persons are more difficult to classify correctly, especially if they are engaged in activities that are relatively informal or intermittent. Many of the changes to the questionnaire were made to deal with such cases. This was accomplished by rewording and adding questions to conform more precisely to the official definitions, making the questions easier to understand and answer, minimizing reliance on volunteered responses, revising response categories, and taking advantage of the benefits of an automated interview. Areas affected by these improvements include:

1. On layoff. Persons on layoff are defined as those who are separated from a job to which they are awaiting recall. The old questionnaire, however, was not structured to consistently obtain information on the expectation of recall. In order to measure layoffs more accurately, questions were added to determine if people reported to be on layoff did in fact have an expectation of recall—that is, had they been given a specific date to return to work or, at least, had they been given an indication that they would be recalled within the next 6 months.
2. Jobsearch methods. To allow interviewers to better distinguish between active and passive methods, the response categories for jobsearch methods were expanded and reformatted. Also, the basic question on jobsearch methods was reworded and followup questions were added to encourage respondents to report all types of jobsearch activity.
3. Hours at work. To improve the accuracy of these data, the series of questions on hours worked was reordered to incorporate a recall strategy that asks for usual hours first, then about possible time taken off or extra hours worked during the reference week, and finally about hours actually worked.
4. Reasons for working part time. Persons who work part time do so either for noneconomic reasons (that is, because of personal constraints or preferences) or for economic reasons (that is, because of business-related constraints such as slack work or the lack of full-time opportunities). Because respondents typically are not familiar with this distinction, the question was reworded to provide examples of the two types of reasons. More importantly, the measurement of working part time involuntarily (or for economic reasons) was modified to better reflect the concept. Starting in 1994, workers who usually work part time and are working part time involuntarily must want and be available for full-time work.
5. Earnings. With the previous questionnaire, respondents were asked to report their earnings as a weekly amount, even though that may not have been the easiest way for them to recall or report their earnings. In the new version, respondents are asked to report earnings in the timeframe that they find easiest, for example, hourly, weekly, biweekly, monthly, or annual. Weekly earnings are automatically calculated for persons who respond on a basis other than weekly.

New data and definitional changes. The questionnaire redesign also made it possible to collect several types of data regularly for the first time, namely:

1. Multiple jobholding. Employed persons now are asked each month whether they had more than one job. This allows BLS to produce estimates of multiple jobholding on a monthly basis, rather than having to derive them through special, periodic supplements.
2. Usual hours. All employed persons are asked each month about the hours they usually work. Previously, information on usual hours was collected from just one-quarter of wage and salary workers each month.
3. Other definitional changes. In addition, several labor force definitions were modified. The most important definitional changes concerned discouraged workers. The Levitan Commission had criticized the former definition because it was based on a subjective desire for work and on somewhat arbitrary assumptions about an individual’s availability to take a job. As a result of the redesign, two requirements were added: For persons to qualify as discouraged, they must have engaged in some jobsearch within the past year (or since they last worked, if they worked within the past year), and they must be currently available to take a job. (Formerly, availability was inferred from responses to other questions; now, there is a direct question.) Also, beginning in January 1994, questions on this subject are asked of the full CPS sample, permitting estimates of the number of discouraged workers to be published monthly (rather than quarterly).

Another important definitional change concerned unemployed persons who were not working just before their jobsearch commenced, that is, new entrants or reentrants to the labor force. Prior to 1994, new entrants were defined as jobseekers who had never worked at a full-time job lasting 2 weeks or longer; reentrants were defined as jobseekers who had held a full-time job for at least 2 weeks and then had spent some time out of the labor force prior to their most recent period of jobsearch. These definitions were modified to encompass any type of job, not just a full-time job of at least 2 weeks’ duration. Thus, new entrants now are defined as jobseekers who have never worked at all, and reentrants are jobseekers who have worked before, but not immediately prior to their jobsearch.

Next: Changes Introduced in 2003