employment and unemployment and indicating trends therein; and

2. Conduct or provide (through contract with institutions, organizations, and individuals, or appropriate Federal or State agencies, or otherwise) for such studies, hearings, research, or other activities as it deems necessary to enable it to formulate appropriate recommendations.

The Commission, according to the Act, is to report its findings and recommendations to the President and the Congress within 18 months after the date of appointment of the first five (of the nine) members of the Commission. Thus the report is due in the fall of 1979.

In preparing its report, the Commission shall consider the following:

1. The methodology of collection of employment and unemployment data at all levels, including National, regional, State, and local statistics; 2. The methodology of analysis of such data, including its relevance and application;

3. The methodology of presentation of employment and unemployment statistics, including the dissemination, current utilization, and application of such statistics;

4. Alternative methods of such collection, analysis, and presentation;

5. The need for additional special statistical surveys (including longitudinal studies) and reports on a continuing basis;

6. The need for additional data and analysis on job vacancies, job turnover, job matching, discouraged workers, part-time workers, youth, minorities, women, and other labor force participants;

7. Accuracy and uniformity of seasonal adjustments in all categories of labor force statistics;

8. Methods of achieving current, accurate, and uniform employment and unemployment statistics on a State and local basis, including their use as a determinant of the allocation of Federal assistance;

9. The need for, and methods to obtain, data relating employment status and earnings, economic hardship, and family support obligations;

10. The extent to which employment and earnings

data assist in determining the impact of public programs and policies upon persons who are economically disadvantaged, unemployed, or underemployed;

11. The availability of and need for periodic information on education and training enrollments and completions in the public and private sectors;

12. Statistical indicators of the relationship between education and training and subsequent employment, earnings, and unemployment experience;

13. The value and usefulness of other statistics regarding employment and unemployment, such as those obtained through operation of the unemployment insurance system;

14. The availability of and need for current and projected occupational information, particularly on a local basis, to assist youths and adults in making training, education, and career choices; and

15. Such other matters as the Commission deems appropriate or necessary, including such matters as are suggested by the President or by the Congress that the Commission deems appropriate.

Data Requirements

To meet the needs for information about labor, there should be a comprehensive system of labor statistics which includes an integrated information set at the national, regional, State, and local area levels. Such an integrated data set should provide information on the labor force and its composition, the number of workers employed by occupation, industry, and area, and the number of hours worked and hours paid for and wages paid for that work together with the benefits received in addition to wages, and the expenditures made by employers for wages and benefits combined. In addition, a comprehensive system should include information about the annual earnings of workers from all employers and from each industry in which they were employed, the prevalence of collective bargaining, wages, earnings, and benefits as well as on work stoppages and labor turnover. Data about the utilization of labor without information about the output of labor is incomplete. Information about labor output and the unit labor cost of producing the goods and services are necessary to an overall evaluation of the economic well-being of the country. Furthermore, as output can be increased without greater efforts on the

part of labor, data are also needed on total factor output and total factor cost.

Core Programs

1. The principal labor statistics programs conducted by the Department of Labor's Bureau of Labor Statistics (BLS) and by the Employment and Training Administration are listed and described in:

a. Major Programs/Bureau of Labor Statistics. The latest published edition is for 1977. The methodology used by the Bureau's publication of these data is also described in BLS Handbook of Methods, Bulletin 1910.

b. Summary of Employment Security Statistical Reports. The latest published edition is for 1977. This publication describes the programs, uses, user, and publications.

2. Data on agricultural employment, wages and hours by State and by type of work are compiled quarterly by the Department of Agriculture's Economics, Statistics, and Cooperatives Service. The data are published in its report Farm Labor which also contains a statement on the source and reliability of estimates. Annual data for hired farmworkers on earnings and days worked, demographic characteristics, and migrant-nonmigrant status. are published in Hired Farm Working Force. 3. Information about Federal Government employment and payrolls is compiled by the Civil Service Commission and published each month in its monthly release of Federal Civilian Manpower Statistics. More detailed information is also published in the Commission's Annual Report, Pay Structures of the Federal Civil Service, and other Commission reports. The Commission also annually publishes data, based on a Federal-State survey, on salaries in 31 administrative, professional or technical occupational categories in State governments. These data are published in the Commission's State Salary Survey, the most recent edition of which provides data as of August 1, 1975.

4. The Department of Defense annually publishes data in its Selected Manpower Statistics of the Department of Defense showing the number of active-duty military, civilian, and reserve forces by branch of service. Monthly summaries are also published.

5. The Bureau of Economic Analysis, Department of Commerce, provides several series of

national and regional information on employee compensation, transfer payments, and employment as part of the national and regional economic accounts.

6. The National Science Foundation conducts studies (through the Bureau of the Census) of manpower characteristics (employment, professional characteristics, and work activities) of scientists and engineers. Data are published in Science Resource Studies Highlights and in the NSF series Surveys of Science Resources. It also compiles and maintains a register of recipients of doctoral degrees in all fields. The data so compiled are published, in statistical form, from time to time.

7. The Maritime Administration, Department of Commerce, compiles and publishes monthly data on areas of employment and maritime manpower in its Merchant Marine Data Sheet. 8. The National Center for Education Statistics, Department of Health, Education, and Welfare, compiles and annually publishes data on the number of teachers at the elementary, secondary, and higher education levels and staffing in local public school systems of various sizes. Every 5 years, the Center provides like information on nonpublic elementary and secondary schools, and on occasion provides data on the number and characteristics of employees in institutions of higher education. 9. The Law Enforcement Assistance Administration, Department of Justice, annually publishes employment data for the policy, judicial, prosecution, legal defense and corrections sectors of the criminal justice system for State, local, and Federal levels of government in Expenditure and Employment Data for the Criminal Justice System.

10. The Federal Bureau of Investigation, Department of Justice, publishes data on law enforcement employees by agency and population size group in its Annual Crime Report. 11. The Labor-Management Services Administration (LMSA), Department of Labor, compiles and publishes data on union financial statistics, and on the characteristics of welfare and pension plans. (The BLS also has programs in this area which complement those of LMSA and of SSA.)

12. The Social Security Administration (SSA) periodically publishes data on the annual earnings of workers covered by Social Security by area in which major earnings were received.

SSA also makes public-use tapes available that permit more detailed analysis of the earnings data, by age, race, sex, industry, locality and quarters of earnings. SSA also compiles and publishes data on employment covered by various related income maintenance and related programs such as on income-loss protection against sickness, private health insurance and employee benefit plans in addition to information directly relating to the Social Security system.

13. The Equal Employment Opportunity Commission compiles and publishes data on the number of women, and other minority employees by broad job category, industry (including State and local governments), and locality.

14. The Interstate Commerce Commission compiles and publishes employment statistics for rail carriers, motor carriers, water carriers, oil pipelines and others covered by the Interstate Commerce Act in its annual publication Transport Statistics in the United States. In addition, ICC publishes employment information in its quarterly Transport Economics.

15. The Railroad Retirement Board compiles and publishes data on characteristics of employees and of beneficiaries in the Railroad Retirement system. Data published annually in Service and Taxable Compensation of Railroad Employees includes summaries of employees' service and taxable compensation in Railroad Retirement Act covered service, number of employees by class of employer, occupation and months of service in a single year.

16. The Bureau of the Census publishes a significant quantity of labor statistics as a result of its work with the Current Population Survey and as important parts of its publications on the decennial census of population and housing and the quinquennial censuses of agriculture, business, and governments. In addition to hard copy publications, the Bureau of the Census also makes data available through public-use computer tapes.

Data Issues

One of the Federal Government's primary concerns is the assessing of the Nation's economy. Since our economy is a dynamic one, with constant changes occurring in the labor market and industry mix, the data requirements of the Federal

Government are also constantly changing and need to be reexamined. This is particularly true as new training and assistance programs are mounted or phased down and as the economic environment changes. This means that periodically a gap develops in the Federal Government's data system. Some of these gaps are discussed below. The existence of a gap does not, however, automatically mean that the Federal Government should mount a data collection program. If the statistical matters discussed below are determined to be necessary to Federal program or policy then a specific justification in terms of the program or policy purpose to be served will be required in the normal course of program and budget review before any work should be undertaken to fill the perceived gap.

One of the most pervasive problems is the lack of an integrated comprehensive data set based on the same sample frames in the same areas that would allow more integrated analysis. This is not to say that data for each series automatically needs to be obtained for each area covered by every other series at the same time. It does mean that if employment hours and earnings surveys (hereafter referred to as the base program) are conducted so as to obtain area level data in "x" areas then other surveys designed to obtain area level detail should be conducted in only those areas that are included in the base program. Unless there is a specific programmatic requirement for data in an area not covered by the base program, all other study programs should then be conducted with reference to cities or areas already included in the base and in each subsequent survey program so that the fullest range of data will be available for the same areas. This observation should not be construed to downgrade the need for better and more comprehensive national level data. However, improvements in data for States and local areas can result in improved national level data. Moreover, the need for more comprehensive data at the national level is such that if resources are not available to improve subnational data, then to the extent resources are available, the national level data should be made more comprehensive and integrated.

The second most pervasive problem is the absence of a comprehensive integrated set of statistics on wages and benefits. Integrated data on the distribution of all workers by demographic characteristics, industry, area, occupation, wage rates, hours worked and hours paid for, weekly earnings, and benefits are not available from any source. Such data are needed for a multitude of purposes. If resources are available, and collection is feasible, data collected on the basis of a sufficiently large sample of employers would provide information about trends and levels that might

replace virtually all industry, area PATC (Professional, Administrative, Technical, and Clerical) wage surveys, and studies of the structure and level of compensation. Moreover, such an integrated study, based on a large annual sample (relating to say June or September) to obtain levels, and smaller samples in the other quarters to collect trend data, would provide the basis of an analytic evaluation of both the level of benefits and their cost. Such data would be useful for determining if further public policy action is required in relation to benefit

areas.

Labor Force, Employment and Unemployment

There is a need for detailed labor force data for each State, Standard Metropolitan Statistical Area (SMSA), and for each of the remaining counties in each State for use in allocating Federal funds under various laws including the Comprehensive Employment and Training Act and the Local Public Works Employment Act (the second title of which also provides for General Revenue Sharing allocations on the basis of a formula the principle component of which is the unemployment rate), for use in program evaluations and for economic analysis and public policy formulation at the national, State, and local governmental levels. The Current Population Survey (CPS) as presently designed does not and will not provide data for all counties or even all SMSA's. There are other data sources and techniques, subsequently discussed, which can serve this need.

Differences in the data on insured unemployment from the Unemployment Insurance (U.I.) system, and the CPS count of the unemployed need to be better rationalized. The addition of questions on the CPS about the receipt of unemployment benefits should be considered for those States where there appears to be a major difference between the CPS estimate of unemployment and the estimate derived on the basis of the administrative records of the State Employment Security Agencies. Such an approach may, however, affect the response rates to and the reliability of the CPS. Hence, testing of this approach through a methods test panel which has been funded in the President's fiscal year 1978 budget should be undertaken before questions about receipt of unemployment benefits are included in the CPS. Also, consideration should be given to matching studies of CPS respondents and recipients of unemployment benefits.

Further, there has been considerable discussion of the perceived need for monthly data and better data by area on discouraged workers. Who are they, when

did they last work, when did they last seek work, and why have they stopped looking for work? Only limited data on discouraged workers once a quarter for the Nation as a whole are now available. The addition of questions about discouragement to the monthly CPS, however, requires careful testing to assure that the basic CPS would not suffer because of either declining response or improper classification of labor force status because of the addition of questions on discouragement. In addition, other discussions have focused on the perceived need for more and better data on the underemployed, by area, and by type of underemployment. For example, do they desire full-time work but can only find part-time employment, or are they underemployed for some combination of these and other reasons (such as working below their skill capacity)? However, this cannot be done until the problems of appropriately defining underemployment are resolved. More information also appears to be needed about the labor force participation of minority groups by area. Many of these gaps could be filled by an expansion of the CPS sample size or by adding some questions to either the monthly core questionnaire, or utilizing a supplementary questionnaire during one or more months. The problems noted above are within the charter for the new National Commission on Employment and Unemployment which, hopefully, will provide guidance in these areas.

In addition the CPS frame from which the sample is selected needs to be continually reevaluted so as to take into account sharp population shifts within an area because of the construction, conversion, or demolition of housing units, including placement of mobile homes, between censuses. It should be noted that the Bureau of the Census does take these factors into consideration in the conduct of the CPS. However, some sharp demographic shifts within areas may not be fully accounted for and the sharp population shifts in some areas that can arise because of the opening, expansion, or closing of a mobile home park, for example, may result in important local area changes that, without continued monitoring may result in samples and results not representative of the area as a whole. Hence, reevaluations of the sample frame should be strengthened and should continue to be of very high priority.

While the CPS sample could be expanded to provide county level labor force estimates, the cost of collecting reliable data for each county would be prohibitive. Moreover, unemployment insurance records tabulated by county of residence, and benchmarked to higher geographic level CPS controls do provide the basis for county level labor force estimates. However, not all administrative rec

ord data are processed by county of residence (the CPS concept), and not all of those receiving Unemployment Insurance (U.I.) benefits are unemployed (some receive wages that are within the "disregard" level contained in virtually all State laws). These administrative records should be adapted for use in the statistical construction of estimates of the labor force and of unemployment consistent with the CPS definitions. This process has begun by BLS and it is hoped that it will shortly be completed. In addition, analytic techniques including regression estimation methods, which could utilize both administrative and survey data, are being and should continue to be thoroughly investigated. To utilize these records in developing labor force estimates, monthly data are also needed on the number of persons employed.

It should be noted that the BLS has recently begun a program of expanding its local area data base of unemployment estimates in order to be able to provide monthly data for each State and each county as needed for program purposes. Some of these data, of course, do not meet the Bureau's usual standards of quality in that they are based on administrative records for very small areas. There are many factors which may contribute to errors in the resulting estimates. Nevertheless, the estimates are based on a conceptually consistent framework and on a conceptually consistent methodology and thus are the best estimates now available. Data at the State level and for the largest SMSA's are published monthly in Employment and Earnings. Data for all other areas are made available through the National Technical Information Service with appropriate

caveats.

The BLS 790 programs do not obtain reliable monthly employment data for all industries or for all SMSA's. Consideration should therefore be given to expanding the program to obtain monthly industrial coverage and to represent each SMSA monthly. The labor force estimates derived from the U.I. and other data should be analyzed regularly to determine factors, such as multiple job holdings, that cause these data to differ from the CPS estimates. This issue (like most of the others in labor force statistics) is also within the charter for the new National Commission.

Another aspect of labor force data that should be expanded is the information about the unemployed who exhaust their unemployment insurance benefits. More information on a more timely basis is needed about the demographic characteristics of the exhaustees, their prior labor force attachment and their future labor force intentions.

Because of multiple job holding the data collected in the BLS 790 program do not provide an unduplicated count of the employed population. Further they do not provide a count of jobs because two persons may hold the same job during the same payroll period. Moreover, the 790 data does not provide counts (estimates) of full-time and of parttime workers. Consideration should be given to the technical feasibility and to the practical utility of obtaining such data. It seems clear that the 790 program, being based on establishment reports, is not the appropriate vehicle with which to attempt to obtain an unduplicated estimate of the employed work force. Such estimates are available from the CPS. However, the estimates from the two series are based on different concepts and statistical techniques and frequently differ considerably in magnitudeeven after reconciliation. Work should be undertaken-perhaps by the new National Commission-towards reconciling the differing concepts, if possible. Further, and even if the estimates cannot be reconciled to sampling variability, the differences between the 790 and CPS estimates should be explained monthly in BLS publications.

The total number of unfilled jobs (i.e., job openings) by industry, occupation, area and offered wages is not now available from any source. It should be noted that some data are available through the Employment Security Automated Reporting System (ESARS) operated by the State Employment Security Agencies (SESA's) in cooperation with the U.S. Employment and Training Administration. Moreover, SESA's throughout the country are producing publications in this area. The SESA data, however, are limited by the lack of complete penetration into the job market. They appear to be particularly deficient in the white collar, trade, service, and government areas. Notwithstanding the gap in knowledge that does exist, serious attention should be given to whether a general-purpose statistical program is best suited for this purpose. Without all of the data elements noted above, a truncated data set would only provide a gross barometer of the changing availability of employment opportunities. Such a barometer is already available from The Conference Board's Index of Help-Wanted Advertising in Newspapers. Moreover, it has been estimated that a job vacancy survey providing all of the data elements noted above-except offered wages, but providing broad regional data instead of data pertaining to job markets would cost between $20 and $50 million per year. If for program purposes the U. S. Employment Service (USES) local offices need better data on un