Short History of the CPS

This document has three parts: first, a short overview of the origins of the CPS; second, a summarized history of the CPS starting with the most recent changes and working back; and third, a more detailed history in chronological order. There are links from capsule statements in the second part into the corresponding detailed descriptions in the third part.

This document is based on Chapter II of Census Technical Paper Number 63. Bibliographical references for all Parts are at the end of Part 3.

PART 1

The Current Population Survey had its origin in a program set up to provide direct measurement of unemployment each month on the basis of a random sample of the U.S. population. There were several earlier attempts to estimate the number of unemployed, using various devices, ranging from guesses to enumerative counts. The problem became especially pressing during the Economic Depression of the 1930's.

This program, the Enumerative Check Census, taken as a part of the 1937 unemployment registration, was the first attempt to estimate unemployment on a nationwide basis using probability sampling. During the latter half of the 1930's, the research staff of the Work Projects Administration (WPA -- known prior to 1939 as the Works Progress Administration) began developing techniques for measuring unemployment, first on a local-area basis and subsequently on a national basis. This research and the experience with the Enumerative Check Census led to the Sample Survey of Unemployment which was started in March 1940 as a monthly activity by the WPA.

In August 1942, responsibility for the Sample Survey of Unemployment was transferred to the Bureau of the Census, and in October 1943, the sample was thoroughly revised. At that time, the use of probability sampling was expanded to cover the entire sample, and new sampling theory and principles were developed and applied to increase the efficiency of the design. The households in the revised sample were in 68 "Primary Sampling Units" (PSU's), all together comprising 125 counties and independent cities. By 1945, about 25,000 housing units were designated for the sample, of which about 21,000 contained interviewed households.

One of the most important substantive changes in the CPS sample design took place in 1954 when, for the same total budget, the sample of PSU's was expanded from 68 to 230, without change in the number of sample households. This redesigned sample, as a result of a more efficient system of field organization and supervision, made it possible to provide more information per unit of cost, and, thus, increase the accuracy of published statistics, and also made it possible to provide more reliable regional as well as national estimates for some characteristics.

Since that time, a major revision in the sample has followed each of the decennial censuses. These and other important modifications in the Current Population Survey are included in the following chronological list.

PART 2

Major Changes in the Survey: A Summary

April 1994 Start of phase-in of new sample.
January 1994 Introduction of a new questionnaire, and of CAPI.
April 1987 Introduction of CATI cases into CPS monthly estimates.
June 1985 Opening of first CATI facility.
January 1985 Changes, mainly to the estimation procedures.
September 1984 Addition of items.
April 1984 Decennial shift of sample base.
January 1983 Inclusion of some Armed Forces personnel.
November 1982 Extension of earnings estimates.
January 1982 Redefinition of race categories.
May 1981 Reduction of sample.
February/March 1981 Introduction of part of the "E" sample into official estimates.
January 1980 Initiation of the use of the "D" sample in official estimates.
January 1980 Initiation of the use of the "B" sample in official estimates.
September/October 1979 Recommendations for further changes.
January 1979 Introduction of methodolgy, and addition of items.
October 1978 Revision of "Control Card" procedures.
September 1975 Introduction of State supplementary samples.
January 1974 Use of inflation-deflation method for deriving independent estimates of population.
December 1971-March 1973 Expansion of sample and updating of input data.
January 1971 and January 1972 Introductionof 1970 census occupational classification.
March 1968 Introduction of more age/sex ratio estimation cells.
January 1967 Change in the concepts of employment and unemployment.
January 1967 Expansion to 449-PSU sample.
January 1963 Increase in descriptive information.
December 1961 Beginning of updating of inputs used in ratio estimation.
October 1961 Conversion to the FOSDIC system.
January 1960 Addition of Alaska and Hawaii.
July 1959 Transfer of functions.
June 1957 Introduction of seasonal adjustment.
January 1957 Change in employment status definition.
May 1956 Expansion to 330-area sample.
July 1955 Changes in survey week.
May 1955 Addition of monthly questions concerning part-time workers.
February 1954 Changeover to 230-area sample.
September 1953 Conversion of tabulations to high-speed electronic equipment.
January 1953 Shift to 1950 census data for ratio estimates.
February 1952 Introduction of document sensing.
July 1949 Introduction of special dwelling places.
August 1947 Revision in sample selection method.
July 1945 Revision of CPS questionnaire.

PART 3

Major Changes in the Survey: A Detailed Chronology

July 1945 -- Revision of CPS questionnaire: The questionnaire was revised to introduce four basic employment status questions. Before that time, the schedule did not contain specific question wording. Special studies showed that this and other defects resulted in the exclusion from the labor force statistics of large numbers of part-time and intermittent workers, particularly unpaid family workers. The wording of these four questions has been modified slightly since 1945, but the basic content has been unchanged.

August 1947 -- Revision in sample selection method: The method of selecting sample units within a sample area was changed so that each selected unit would have the same basic weight in the tabulations. This change simplified tabulations and estimation procedures.

July 1949 -- Introduction of special dwelling places: The sample coverage was extended to special dwelling places, such as hotels, motels, and trailer camps. This change led to improvements in the statistics, because residents of such places have characteristics different from those of the remainder of the population.

February 1952 -- Introduction of document sensing: The form used to record the answers to the CPS questionnaire was converted to a document-sensing card. In this procedure (replaced in 1961 by the FOSDIC system), entries were made by drawing a line through the oval representing the correct answer, using a special pencil with electographic lead. Punchcards were automatically prepared from these answer-recording cards by a special documenting-sensing machine.

January 1953 -- Shift to 1950 census data for ratio estimates: Starting in January 1953, population data from the 1950 census were introduced into the Current Population Survey estimation procedure. (See Chapter V of Census Technical Paper Number 63 for a description of these estimates.) Prior to that date, the ratio estimates had been based on the relationships calculated from the 1940 census for the first-stage ratio estimate and on the population data from the 1940 census, brought forward to take account of recorded changes such as births and deaths, for the second-stage ratio estimate. In September 1953, "color" was substituted for "veteran status" in the second-stage ratio estimate, making it feasible to publish some separate, absolute numbers for persons by race, whereas prior to this change only percent distributions had been provided.

July 1953 -- Change to 4-8-4 rotation system: The present sample rotation system was adopted, under which households are interviewed for 4 consecutive months, leave the sample for the next 8 months, and return for 4 consecutive months, which are therefore the same four calendar months as the first four, but a year later. Prior to that time, households were interviewed for 6 months and then replaced. The new system provided some year-to-year overlap in the sample, thus improving the measurement of the statistics over times. (See Chapter III of Census Technical Paper Number 63.)

September 1953 -- Conversion of tabulations to high-speed electronic equipment: This change sped up the tabulating considerably, made possible improvements in estimation methods, and substantially expanded the scope and content of the tabulations for basic data and computation of sampling variability. A shift to more moderncomputers was made in 1959; this process has continued as equipment is updated and replaced.

February 1954 -- Changeover to 230-area sample: The CPS sample was expanded from 68 to 230 sample PSU's, although the overall sample size of 25,000 designated sample units was retained. The 230 areas comprised 453 counties and independent cities. At the same time, a substantially improved estimation procedure (composite estimate) was introduced, which took advantage of the large overlap in the sample from month to month. These two changes improved the reliability of most of the major statistics by an amount equivalent to that which would have been obtained by doubling the sample size.

May 1955 -- Addition of monthly questions concerning part-time workers: Monthly questions on the reasons for part-time work were added to the standard set of employment status items. This information had been collected quarterly (or even less frequently) in the past and was found to be valuable in studying current labor market trends.

July 1955 -- Changes in survey week: The CPS survey week was changed to the calendar week containing the 12th day of the month, for greater consistency with the time reference of other statistics in the employment field. Previously, the survey week had been the calendar week containing the eighth day of the month.

May 1956 -- Expansion to 330-area sample: The CPS was expanded from a 230-PSU to a 330-PSU sample. The overall sample size was increased by roughly two thirds to a total of about 40,000 designated units (about 35,000 occupied units). The expanded sample was located in 638 counties and independent cities, with at least some households in each of the 48 contiguous States. All of the former 230 areas were continued in the expanded sample. The expansion increased the reliability of the major statistics by around 20 percent and made possible publication of greater detail.

January 1957 -- Change in employment status definition: Two relatively small groups of persons formerly classified as employed "with a job but not at work" were assigned to different classifications as a result of a comprehensive interagency review of the Goverment's employment and unemployment data. These groups were (1) persons on layoff with definite instruction to return to work within 30 days of the layoff date and (2) persons waiting to start new wage-and-salary jobs within 30 days of interview. Most of the persons in these two groups were shifted to the unemployed classification. The only exception was the small subgroup in school during the survey week and waiting to start new jobs; it was transferred to "not in labor force." The changes in definition did not affect the basic question or enumeration procedures.

June 1957 -- Introduction of seasonal adjustment: A limited amount of seasonally-adjusted data on unemployment were introduced early in 1955. Some extension of the data, using more refined seasonal adjustment methods programmed on electronic computers, was instituted in July 1957, including a seasonally-adjusted rate of unemployment and charting of seasonally-adjusted total employment and unemployment. Significant improvements in methodolgy grew out of research conducted at the Bureau of Labor Statistics (BLS) and at the Bureau of the Census in the ensuing years [7].

July 1959 -- Transfer of functions: Responsibility for the planning, analysis,and publication of the labor force from the Current Population Survey was transferred to the Bureau of Labor Statistics as part of a major exchange of statistical functions between the Commerce and Labor Departments. The Bureau of the Census continued to have (and still has) responsibility for the collection and tabulation of these statistics, for maintenance of the CPS sample, and for related methodological research. Interagency review of policy and technical issues continues to be conducted by committees under the aegis of the Statistical Policy Division, Office of Management and Budget.

January 1960 -- Addition of Alaska and Hawaii: Upon achieving statehood, Alaska and Hawaii were introduced into the independent estimates of the population and into the sample survey, thereby increasing the number of PSU's in the sample from 330 to 333. The addition of these two States affected the comparability of National population and labor force data with those from previous years. This inclusion resulted in an increase of about half a million in the noninstitutional population of working age and of about 300,000 in the labor force, four fifths of this latter increase being in nonagricultural employment. The levels of other labor force categories were not appreciably changed.

October 1961 -- Conversion to the FOSDIC system: The form used for recording the answers to the CPS questionnaire was changed to the "FOSDIC" type used in the 1960 census. (This system remained in use until the introduction in January 1994, as part of the CPS Modernization effort, of computer-aided interviewing.) Entries were made on a FOSDIC form by filling small circles with an ordinary lead pencil. Microfilms of these answer forms were scanned by a special mechanical reading device which transferred the information directly to computer tape. This system permitted a physically larger form and a more flexible arrangement of items than did the previous document-sensing procedure, and it did not require the preparation of punchcards.

December 1961 -- Beginning of updating of inputs used in ratio estimation: During the period from December 1961 to March 1963 inclusive, the CPS sample was revised gradually. This change was made to reflect the changes in population size and distribution, and in the industrial mix among areas, revealed by the 1960 census. The overall sample size was unchanged, but the number of sample areas was increased slightly to 357 PSU's, to provide for greater coverage in fast-growing portions of the country. Also, in a major part of the sample, selection of units from census lists, developed in the 1960 census, was introduced to replace area sampling [8, p.69]. These changes resulted in a further gain in reliability of about 5 percent for most statistics. The use of updated population information from the census was introduced in April 1962 into first- and second-stage ratio estimates used in the CPS.

January 1963 -- Increase in descriptive information: In response to recommendations of a special review committee [9], two new items were added to the monthly questionnaire. The first was an item, formerly carried only intermittently, on whether the unemployed were seeking full- or part-time work. The second was an expanded item on household relationship, formerly included only annually, to provide more detail on the unemployed persons by household relationship and marital status.

January 1967 -- Expansion to 449-PSU sample: The CPS was expanded from a 357-PSU to a 449-PSU sample. As a result of an increase in total budget, the overall sample size was increased by roughly 50 percent to a total of about 60,000 housing units (52,500 occupied units). The expanded sample had households in 863 counties and independent cities, with at least some coverage in every State. This expansion increased the reliability of the major statistics by about 20 percent and made possible the publication of greater detail.

January 1967 -- Change in the concepts of employment and unemployment: In line with the basic recommendations of the President's Committee to Appraise Employment and Unemployment Statistics [9], an experimental program was conducted for several years to develop and test proposed changes in the concepts. The principal improvements resulting from this research were put into effect in the household survey in January 1967. The changes included a revised age cutoff in defining the labor force and a slightly revised definition of unemployment; they also added probing questions to improve the information on hours of work, on the duration of unemployment, and on the self-employed. The differences arising in the estimates of level and of month-to-month change due to these revisions in the concepts are relatively small. Further information on the nature of these changes is given in [6].

March 1968 -- Introduction of more age/sex ratio estimation cells: There are three categories for race reported on the CPS questionnaire: white, Negro, and other. After the separate ratio estimates for Negro and for other, a second set of factors, by age and sex, is applied separately, once to the total of all persons with race reported as not white, and once to the total of all persons with race reported as white; this set of factors uses a larger number of age cells than does the first. The procedure previously used did not apply separate factors for Negro and for other. This change amounts essentially to an increase in the number of ratio estimation cells from 68 to 116. (See Chapter V of Census Technical Paper Number 63.)

January 1971 and January 1972 -- Introduction of 1970 census occupational classification: The questions on occupation were made more comparable to those used in the 1970 census by adding a question on major activities or duties on that job. The new classification was introduced into the CPS coding procedures in January 1971. The tabulations were produced in the revised version beginning in January 1972 [10].

December 1971-March 1973 -- Expansion of sample and updating of input data: During the period from December 1971 to March 1973 inclusive, the CPS sample was revised gradually to reflect the changes in population size and distribution shown by the 1970 census. As part of an overall sample optimization, the sample size was reduced slightly (from 60,000 to 58,000 total units designated), but the number of sample areas was increased to 461 PSU's. This new sample was named the "A" sample. As part of this re-design, a change was made from requiring clusters to consist of six nearby (but not contiguous) households to requiring them to consist of only four households, but usually of contiguous ones. This change was instituted after Census Bureau studies indicated that a smaller cluster size would result in a more efficient sample. Thus, even with the reduction in total sample size, there was a small gain in reliability for most characteristics. The noninterview adjustment and the first-stage ratio estimation adjustment were also changed slightly, in order to improve the reliability of estimates for central cities and for the balances of SMSA's.

The independent estimates of the civilian noninstitutional population by age, race, and sex, used for the second-stage ratio estimation procedure, were changed over to the 1970-census base in January 1972.

January 1974 -- Use of inflation-deflation method for deriving independent estimates of population: The derivation of independent estimates of the civilian noninstitutional population by race, sex, and age, used in the second stage of ratio estimation in preparing the monthly labor force estimates, was changed over to the inflation-deflation method. (See Chapter V of Census Technical Paper Number 63.).

September 1975 -- Introduction of State supplementary samples: An additional sample, comprising about 14,000 assigned interviews each month, was introduced beginning in July 1975 to supplement the national sample in 26 States and in the District of Columbia. A total of 165 new PSU's was involved. The supplement was added to meet a specific reliability standard for estimates of the annual average number of unemployed persons for each State. Effective August 1976, because of the introduction of an improved estimation procedure and because of a modification of the reliability requirement, the supplement PSU's were dropped from three States (ch. VI) and the size of the supplemental sample was reduced to about 11,000 assigned interviews in 155 PSU's (See Appendix J of Census Technical Paper Number 63.).

October 1978 -- Revision of "Control Card" procedures: Changes were made to the methods for collecting information about household relationship, marital status, race, and ethnicity. In particular, the response entered for race became determined by the respondent rather than, as previously, by the interviewer.

Earnings questions were added to the version of the questionnaire used with the two outgoing rotation groups (the one then at its fourth month in the sample. called outgoing because it would be out temporarily (for the next eight months), and the one then at its eight month in the sample, called outgoing because it was then leaving the sample permanently). Earnings items added included usual hours worked, hourly wage rate, and usual weekly earnings. Earnings items are asked only of currently employed wage-and-salary workers.

January 1979 -- Introduction of methodolgy, and addition of items:

A new, two-level first-stage ratio estimation procedure was introduced. It was designed to improve the reliability of estimates for areas distinguished as either metropolitan or nonmetropolitan.

Items were added to begin collection of children's transcription data, including relationship, age, sex, race, and origin.

September/October 1979 -- Recommendations for further changes: The final report of the National Commission on Employment and Unemployment Statistics (the NCEUS, commonly known (after its chair) as the Levitan commission) was issued. This set of recommendations shaped many of the changes which have since been made to the CPS.

January 1980 -- Initiation of the use of the "B" sample in official estimates: The "B" sample was constructed so as to provide estimates of publishable quality for the 10 largest States on a monthly basis. Prior to this improvement, estimates for States were guaranteed to be of publishable quality only on an annual-average basis, and only estimates for the Nation were guaranteed to be of publishable quality on a monthly basis. The improvement was achieved by adding about 450 households to the sample, and by redesigning it to extend it to 629 PSU's.

February/March 1981 -- Introduction of part of the "E" sample into official estimates: The "E" sample was originally designed to increase the reliability of sub-State estimates, and included about 6,500 additional assigned units. Only about 100 of these units were ever used in the official CPS estimates.

May 1981 -- Reduction of sample: The sample was reduced by approximately 6,000 units (predominantly from the "B" sample), bringing the total assigned sample size to approximately 72,000 assigned units.

January 1982 -- Redefinition of race categories: The race categories in the second-stage adjustment were changed from White/NonWhite to Black/NonBlack. These changes were made to conform to classification differences in race that existed between the 1980 census and the CPS. The change did not result in notable differences in published household data; however, it did result in more variability for certain "White" and "Black and Other" characteristics.

As is customary, the CPS uses population controls from the most recent decennial census. Beginning in January 1982 these estimates were based on findings from the 1980 census. The use of the 1980-based population controls, in conjunction with the revised second-stage adjustment, resulted in about a 2 percent increase in the estimates for total civilian noninstitutional population 16 years and over, civilian labor force, and unemployed persons. The magnitude of the differences between the controls based on the 1970 census and those based on the 1980 census affected the historical comparability and continuity of major labor force series; therefore, approximately 30,000 series were revised by the BLS back to 1970.

November 1982 -- Extension of earnings estimates: The "earnings" series was extended to include items on union membership and union coverage.

January 1983 -- Inclusion of some Armed Forces personnel: Persons in the Armed Forces stationed in the United States started being included in the national labor-force and employment totals. This enlargement of the labor force by persons who all were employed reduced the national unemployment rate by about 0.2 percent. The occupational and industrial data started being coded using the 1980 classification systems. While the effect on industry-related data was minor, the conversion was viewed as a major break in occupation-related data series. Census developed a "list of conversion factors" to translate occupation descriptions based on the 1970 census coding classification system into their 1980 equivalents. Most of the data historically published for "Black and Other" population groups were replaced by data that relates to the "Black" population.

April 1984 -- Decennial shift of sample base: The use of the 1970 census as the basis of the CPS sample begins to be phased out. This shift to the use of the 1980 census as the basis of the sample will continue through a series of changes that will be completed by July 1995. The redesigned sample uses data from the 1980 census to update the sampling frame; it takes advantage of recent research findings to improve the efficiency and quality of the survey; and it uses a State-based design to improve the estimates for the States without increasing the same sample size.

An indirect effect of this introduction of the new sample was a disruption of sub-national geographical estimates. That is, during the phase-in the old sample reflected 1970 census geography identifiers, while the new sample used those from the 1980 census. The 1980 definitions were converted to 1970-equivalent designations through December 1994. Starting from January 1985, the reverse of this conversion occurred. However, due to coverage differences between the old sample and the new, the BLS suspended publication of metropolitan-nonmetropolitan estimates until the 1980 census sample was fully phased in.

September 1984 -- Addition of items: Collection of veterans' data was initiated for females.

January 1985 -- Changes, mainly to the estimation procedures: Data from the 1980 census and the new sample began to be used, in conjunction with some new methodologies for the second-stage adjustment. The primary change was that controls for "White", "Black", and "Other" population groups, by sex and age groups, replaced controls for "Black" and "NonBlack" population groups. In addition, a separate, intermediate step was added to control to the Hispanic population. The combined effect of these changes on labor force estimates and aggregates for most population groups was negligible; however, the Hispanic population and associated labor force estimates were greatly affected. As a result, revisions to the affected Hispanic estimates were made back to January 1980, to the extent possible.

June 1985 -- Opening of first CATI facility: A site for conducting Computer-Assisted Telephone Interviewing was established at Hagerstown, Maryland. A series of tests was conducted over the next few years to identify and resolve the operational issues associated with the use of CATI. Later tests focused on CATI-related technical issues such as data quality, costs, and mode effects on labor force estimates. Samples used in these tests were completely separate from the CPS.

April 1987 -- Introduction of CATI cases into CPS monthly estimates: Initially, CATI started with 300 cases a month. As operational issues were resolved and as new telephone centers were opened (at Tucson, Arizona in May 1992, and at Jeffersonville, Indiana in September 1994), the CATI workload was gradually increased to about 9,200 cases a month by January 1995.

January 1994 -- Introduction of a new questionnaire, and of CAPI: After much research and testing, a redesigned questionnaire was introduced for the development of official CPS estimates. This is the single most significant change in the history of the CPS.

Prior to this redesign, the survey questionnaire had been unchanged for almost three decades. The redesign of the questionnaire had four main objectives: (1) to measure the official labor force concepts more precisely; (2) to expand the amount of data available from the CPS; (3) to implement several changes in concepts and definitions; and (4) to make the best possible use of the potential advantages of a computer-assisted interviewing environment which was to be introduced at the same time. In particular, certain data would be collected more frequently, permitting the development monthly, rather than annually, of estimates of multiple job-holding; and certain data would be collected more extensively, permitting the development of estimates of usual hours worked on the basis of the full sample rather than of only one quarter of it. Further, the wording of questions was changed to reflect modifications in the definitions of several labor-force concepts: discouraged workers, unemployment, new entrants and re-entrants, and full- and part-time workers.

The difficulties naturally incident upon the introduction of a longer and more complicated questionnaire were eased by the simultaneous introduction of Computer-Assisted Personal Interviewing. The CPS interviewers in the field started carrying lap-top computers equipped with a knowledgeable, interactive version of the new questionnaire. Computerization allows for the use of a very complex questionnaire, incorporating many contingent paths through many contingently appropriate subsets of the entire body of available questions, without increasing the burden of time and attention upon the respondent or the interviewer. It increases consistency by reducing interviewer error through the use of probes and of editing at time of interviewing, and it permits dependent interviewing -- information reported in one month (industry/occupation, retired/disabled statuses, and duration of unemployment) can be confirmed or updated in subsequent months without the need for asking a long series of questions and entering the responses from scratch.

Also at this time, the independent population controls used in the second-stage estimation procedure underwent their usual decennial updating, this time to ones derived from the 1990 census. For the first time, however, the controls reflected a measurement of the estimated Census undercount.

April 1994 -- Start of phase-in of new sample: A phase-in of the redesigned sample, based on the 1990 Census, begins at this time. It will be completed in 16 months, because the mechanism is that each month the incoming rotation group new to the sample will have been selected under the redesign, whereas the rotation group which it replaces, the one which in the previous month was experiencing its final month in the sample, had been selected under the old design. Therefore, the 4-8-4 in-sample pattern causes this group-by-group replacement process to take 16 months.

References

1. Eckler, A. Ross. The Bureau of the Census. New York: Praeger, 1972

2. Frankel, Lester R., and Stock, J. Stevens. "On the Sample Survey of Unemployment." Journal of the American Statistical Association 37 (March 1942): 7780.

3. Hansen, Morris H. and Hurwitz, William N. "On the Theory of Sampling From Finite Populations." Annals of Mathematical Statistics XIV (1943): 333362.

4. Hansen, Morris H, Hurwitz, William N. and Madow, William G. Sample Survey Methods and Theory Vol. I, Methods and Applications. New York: John Wiley and sons, 1953.

5. Hansen, Morris H., Hurwitz, William N., Nisselson, Harold, and Steinberg, Joseph. "The Redesign of the Census Current Population Survey." Journal of the American Statistical Association 50 (September 1955): 701719.

6. U.S. Department of Commerce, Bureau of the Census. The X11 Variant of the Census Method II Seasonal Adjustment Program, by Julius Shiskin, Allan H. Young, and J.C. Musgrave. Technical paper 15. Washington, D.C.: Government Printing Office, 1965.

7. U.S. Department of Commerce, Bureau of the Census. The Current Population Survey: A Report on Methodology, by Joseph Steinberg, Thomas B. Jabine, and Leon Pritzker. Technical Paper 7. Washington, D.C.: Government Printing Office, 1963.

8. President's Committee to Appraise Employment and Unemployment Statistics. Measuring Employment and Unemployment. Washington, D.C.: Government Printing Office, 1962.

9. U.S.Department of Commerce, Bureau of the Census, and U.S. Department of Labor, Bureau of Labor Statistics. "Concepts and Methods Used in Labor ForceStatistics Derived From the Current Population Survey," Current Population Reports. Special Studies Ser. P23, No. 62. Washington, D.C.: Government Printing Office, 1976.

10. U.S. Department of Labor, Bureau of Labor Statistics. Employment and Earnings. Vol.17 No. 8. Washington, D.C.; Government Printing Office, 1971.

Basic Monthly Survey History and Concepts Page

CPS Home Page