light on the occupational composition of the employed labor force but would also provide occupational data about the unemployed labor force. In addition, it may be possible to obtain data about the occupational characteristics of those who are then out of the labor force but who had recent employment histories and who-in an emergency or under the right set of socioeconomic circumstancesmight reenter the labor force. Such data elements cannot be obtained through any technique short of a census procedure. An annual but more limited set of occupational data would be provided by a matching of CPS and Social Security files (see next section on compensation).

To meet the national and area data needs it is therefore recommended that the BLS should currently continue providing State Employment Security officials with the necessary guidance in the OES survey including technical manuals, consultation services, sample selection techniques, editing and screening procedures, standardized questionnaires, estimating procedures, and complete processing systems. In addition, the BLS should continue to develop and improve procedures to measure current and projected occupational supply, including replacement needs, occupational mobility, data on the number of unemployed by occupation, and the extent of occupational training by private industry. However, the data collection aspect of the OES program as currently constituted should be terminated when the 1980 Census data are available and the plans for the 1985 mid-decade effort have been sufficiently formalized to assure that the needed national data will be obtained in that effort. The census data will then replace the OES core data collection effort. Nevertheless, the BLS should continue to provide technical guidance and assistance to those States and local areas that desire to conduct an occupational employment survey to meet a specific need not satisfied by the data from the censuses. Such studies, however, should be entirely funded by the jurisdictions conducting them. Moreover, the BLS should continue its projections program and should incorporate data from the censuses and from other available studies into the data base used in making such projections.

Compensation

As previously noted, one of the most pervasive labor statistics problems is the absence of a comprehensive integrated data set on wages and benefits. In addition, government employment, accounting for almost 15% of the employed work force, is seriously underrepresented in the wage and hours data series. The BLS 790 program does not provide earnings and

hours data for the government sector on a monthly basis and should be expanded to do so. The BLS industry and area wage survey programs do not include governments-and its Municipal Government Wage Survey program excludes States, counties, special districts and Federal employment. That program which only includes 27 cities of 500,000 population or more should be revised to cover all areas (about 70) within scope of the area wage survey program, and within those areas should be expanded to represent all units of government including the Federal Government and special governmental units such as education. Moreover, the Professional, Administrative, Technical and Clerical (PATC) surveys should be expanded to cover governments.3

The October 1976 report on State and Local Government Employee Compensation Data Needs by the Council on Wage and Price Stability contains recommendations for expanding Federal collection of data on State and local government employment compensation. The recommendations, in priority order, are:

1. The Bureau of Labor Statistics should expand its Average Weekly Earnings series to include State and local government employees.

2. The Bureau of Labor Statistics should develop a series to report the size of collective bargaining settlements in the State and local sector.

3. The Bureau of Labor Statistics should expand the coverage of its Municipal Wage Surveys to cover a larger portion of the public sector.

4. The Bureau of Labor Statistics should conduct studies of the components of total compensation in the public sector.

5. The Civil Service Commission should expand the scope of its State Salary Surveys to cover benefits.

'In addition, a special program should be instituted to provide for a matching of occupational classifications with common grade or level designations used by Federal, State, and local governments. This effort should be jointly undertaken by the Civil Service Commission, the Advisory Committee on Intergovernmental Relations and the Bureau of Labor Statistics working in concert with the appropriate associations of the States, counties, cities, and other units of government and with the major unions and employee associations involved in State and local government. Further, private sector occupations should also be analyzed and classified into the same grade or level designations for statistical analysis. The results of this total effort would facilitate comparison between governmental units at the same (e.g., State) and different (e.g., city, county, and State) levels of government and would further facilitate comparison of wages and salaries and of total compensation between the public and private sectors of the economy by area (also see the chapter on standards development).

It should be noted, however, that the Council's findings and recommendations indicate that there is "a need" for data but does not indicate what purpose the data needed are intended to serve. Hence, the Council's prioritization of recommendations and the recommendations themselves cannot be assessed in terms of who needs what for what. Moreover, the absence of any statement of purpose (other than "The overall figures are especially useful to those who wish to keep track of trends in public sector earnings...") makes it difficult to assess the relative importance of these data to the Council and to others versus other possible data series pertaining to the private sector.

The industry and area wage survey programs exclude small units (with the size of establishment cutoffs varying by industry). These exclusions are based on the belief that there are too few workers in occupations that match survey definitions in the excluded units to affect wage and hours distributions and mean estimates. Nevertheless, some data (from New York State studies) do exist that suggest that there are enough workers in matching occupations in small firms to markedly affect the wage distributions and that excluding these observations tend to skew the distributions to the right. Accordingly, the size of firm exclusions should be reexamined through surveys that utilize the same collection methodology as do the ongoing studies to test the reasonableness of continuing the size of establishment cutoffs.

It is conceivable that data on straight-time wages, premium payments, and hours worked and hours paid for can be obtained through a household survey such as the CPS. Such an approach can yield powerful analytic tools because of the family and other demographic and human capital information already available in the CPS. However, household data on earnings significantly underreport wage and salary information at lower earnings intervals and earnings by the self-employed (including those with high earnings). Since one of the uses such data would be put to would be an analysis at various income levels, an understatement of incomes would limit the value of the analysis. Therefore, before any full-scale household survey of wage rates is undertaken a validation study of wage rate data collected from households by examining employer records for the same workers is essential. Such an effort is now underway.

Annual wage and salary data are obtained through the March supplement to the CPS and from the Social Security Administration's records. The former series does not have sufficient sample to permit industry analysis and the latter does not contain data on occupation, hours paid or weeks worked,

education and other demographic characteristics beyond age and race. Moreover, data based on the SSA files are available only after several years have elapsed since the end of the reference year, and the SSA data exclude all earnings not covered by social security. One of the major uses of the data is in the evaluation of manpower programs and the omission of these variables and the time in between dates of reference and availability create serious deficiencies for that use. Accordingly, the social security one percent annual employer-employee data file should be expanded so as to provide data on occupation (at least in broad groupings), aggregate hours and weeks worked during the year, education and family characteristics. This might be accomplished by either merging existing data in Bureau of the Census files with SSA data, by conducting periodic sample surveys, or by some combination of these and other techniques. In addition, processing of these data should be expedited. In addition, the CPS data appear to understate earnings particularly at the lower earnings end of the distribution. Some understatement appears to also occur at the highest earnings level. Therefore, if possible, new approaches should be developed to obtain better measures of wage and salary earnings in the CPS.

Employee Benefit Plans

The statistical series on employee benefit plans are woefully inadequate. They do not indicate who is, or how many people by demographic characteristics are covered by a pension, health, or life insurance program by level of benefits nor do they indicate the cost of the program to the employer and to the employee. What is needed is an analytical framework built on representative samples based on establishments that would cross relate level of benefits, number of persons by demographic characteristics that are covered, by industry, area, occupation and earnings level, by cost of plan on a unit base (e.g., hours, persons, etc.). In addition, data on benefit forfeitures, and out-of-pocket expenditures of individuals in addition to those covered in whole or in part by the benefit plan are needed for use in formulating and implementing public policy.

Industrial Relations

The industrial relations statistics are derived from three sources: (1) the BLS biennial survey of national unions and employee associations which provides union membership by industry, State, sex, and selected occupations in the aggregate and for particular unions; (2) annual data from the May supplement to the CPS which provides data on earnings by demographic characteristics for union and

nonunion workers; and (3) BLS establishment surveys which provide estimates of collective bargaining agreement coverage. However, the latter two sources do not provide satisfactory estimates of the number of workers actually covered by collective bargaining by industry, by area, or by union. The data collected in establishment surveys relate to the number of production or nonsupervisory workers (in total) in establishments where a majority of the workers are covered by bargaining agreements.

The surveys are either based on relatively efficient samples of the entire economy which do not permit the derivation of estimates below the industry division level, or relate to a handful of two-, three-or four-digit (SIC) industries and thus exclude the majority of industries at any level of industry classification. Moreover, small establishments are excluded from most of the surveys that obtain such data. The estimates of collective bargaining agreement coverage obtained from the May CPS supplement diverge considerably from estimates obtained from either of the other sources particularly when industry data are compared. This divergence should either be significantly reduced to a difference explainable by sampling variation or, if this is not feasible, the May supplement to the CPS should be dropped since the estimates based on establishment and membership studies are generally considered to be more accurate.

To obtain more comprehensive data on contract coverage with detail by industry, questions about collective bargaining coverage should be added to the 790 program. Such an approach had been previously explored without success. However, the approach should again be tested because if the data could be collected it would also permit analysis of differential average wage and hours patterns by union-nonunion characteristics by industry. Further the addition of such questions would provide a statistical frame with which to select sainples of union contracts covering workers in establishments of all sizes. Data describing the prevalence of contract provisions now excludes all contracts covering fewer than 1,000 workers.

Productivity

The official BLS series on productivity (output per unit of input) relate output to units of labor input only. Since labor input per unit of output can change without increased or decreased effort because of technological changes, among others, new measures to measure output per unit of total factor input should be developed for simultaneous publication with the labor productivity series.

The productivity measures now available are particularly sparse in nonmanufacturing industries. The series should be expanded to cover as many of these industries as conceptually and technically possible.

The measurement of productivity in the governmental sector appears to be conceptually deficient particularly with respect to research, defense, legal, and administrative activities. Moreover, coverage-even in the Federal sector-is incomplete. These problems and other issues are being examined by the National Commission on Productivity Statistics. This commission, composed of eminent economists and statisticians, will complete their work and issue a report late in 1978.

Programs To Be Discontinued

No programs are recommended for total discontinuance at this time. However, at a later time some programs may be integratable and a single program might provide the data now produced by several. For example, the Employment Cost Index (ECI) could subsume and eliminate the need for the adjusted hourly earnings index (which is simply a poor proxy-but the only one now existing for a measure of wage rate change), and the biennial compensation expenditure and payroll hours program could be eliminated if the ECI were expanded so as to provide periodic measures of level in the entire economy (by broad sector) as well as a measure of trend on a fixed weight basis. Such an economy wide level might be obtained in those years that the Bureau reweights the ECI to make it more current in its coverage. Other suggestions for program modification and elimination are discussed in the section on data issues.

Introduction

A common requirement of many social programs and issues is the number of people in a population of interest. New priorities for revenue sharing and block grants require current estimates of the population for extremely small areas. In addition to information about the number of people, data are needed on the geographic distribution of the population; its characteristics-age, sex, race, ethnic origin and other demographic and social attributes-as well as its income, wealth and other economic variables. The requirements for population estimates and projections are not limited to the population of the United States. Our national interests and security require that we have estimates and projections for most other nations as well.

Data are also needed for families and households. The number of families, their size, composition and age structure are important variables for demographic analysis. These kinds of family data are required to plan for and to evaluate various income maintenance, health, education, and similar programs.

Producing Agencies

There are a limited number of agencies involved in the production of basic population data. The greatest responsibility falls on the Bureau of the Census. The Census Bureau is responsible for the constitutionally mandated decennial count as well as the new middecade census requirements. Census is also responsible for providing the official intercensal population estimates. The Department of Commerce directive 13 (formerly OMB Circular A-46 exhibit I) requires that all Federal assistance programs which utilize population as an element in a grant formula must use the most current estimate from the Census Bureau, except where other measures, such as the decennial census, are required by law. In addition the Bureau produces domestic population projections as well as foreign estimates and projections. The Bureau of Economic Analysis also produces population projections which are widely used by Federal programs. Census provides regular data on the

characteristics of the population through the monthly current population survey and its supplements.

It would not be possible for Census to develop accurate estimates of the population without good vital statistics data-births and deaths. The collection of these vital records is the responsibility of the National Center for Health Statistics (NCHS) through a cooperative effort with the States. NCHS also conducts the survey which provides the data for the World Fertility Survey.

The General Revenue Sharing Program requires Census to develop estimates of the population and per capita income for about 39,000 units of government. In order to provide such small area data, Census has had to rely on information from the Internal Revenue Service. Periodically, questions have been included on personal income tax returns (SF-1040) to ascertain the actual political jurisdiction in which the taxpayer lives. These data provide a measure of mobility needed to develop estimates for extremely small areas.

The Center for Population Research of the National Institutes of Health is actively involved in supporting research which sheds light on the dynamics of demographic change. While this research has less direct impact on population statistics, it is useful in anticipating changes in population. It also provides insights into the causes and correlates of population change which frequently has direct policy relevance.

The Immigration and Naturalization Service (INS) of the Department of Justice is developing a program to estimate the number and distribution of illegal aliens in the United States.

Users

The users of domestic population data include virtually all Federal agencies, especially those involved in Federal domestic assistance programs. Most of these programs distribute funds based on a simple population formula or on a formula which uses population as the denominator for some ratio.