January 1994 Employment Situation

Historical release narrative reproduced from the official public-domain source. Statistical tables, standard technical appendices, administrative material and attendance rosters are omitted; formatting is adapted for reading. Figures retain the release's historical context and may have been revised. Consult the linked source for the complete document.

Payroll employment, as measured by the survey of nonfarm business establishments, edged up by 62,000 in January, the Bureau of Labor Statistics of the U.S. Department of Labor reported today. This gain was well below the average for recent months, being held down by the adverse impact of severe weather in much of the country.

The unemployment rate was 6.7 percent in January. This and other measures from the survey of households are the first official estimates produced using a totally redesigned survey. The results from the December and January surveys should not be directly compared. Findings from a test conducted in 1993 showed that joblessness as measured by the new survey questionnaire was, on average, somewhat higher than on the old basis and that many other labor force estimates also were affected. In addition, the household survey data reflect the introduction of revised population estimates based on the 1990 census, as adjusted for the undercount, which markedly raised labor force and employment levels.

Unemployment (Household Survey Data)

The unemployment rate was 6.7 percent in January (seasonally adjusted), as measured using the new questionnaire and survey procedures. The number of unemployed persons was 8.7 million. On the old basis, the rate in December was 6.4 percent and the jobless level was 8.2 million; both had been drifting downward throughout 1993. The jobless rates for the major demographic groups in January were as follows: adult women (6.0 percent), adult men (5.9 percent), teenagers (18.4 percent), whites (5.8 percent), blacks (13.1 percent), and Hispanics (10.6 percent). (See tables A-1 and A-2.)

The measurement and classification of unemployed persons by reason for unemployment also have changed somewhat with the survey redesign. An estimate of the number of persons unemployed because their temporary job ended is now available separately for the first time (not seasonally adjusted). In January--on a seasonally adjusted basis--the grouping "job losers and persons who completed temporary jobs" (which roughly approximates the "job loser" category in the old survey) made up 51 percent of the unemployed; 14 percent of all unemployed persons were on temporary layoff (expecting recall). (See table A-6.)

The proportion of unemployed persons who were reentrants to the labor force (33 percent) was markedly higher based on the new figures, while the proportion who were new entrants was lower (7-1/2 percent). Much of this shift reflects the redefinition of reentrants in the new survey, whereby persons are no longer required to have at least 2 weeks of full-time work experience to be classified as reentrants; any work experience, including only part-time jobs, will now suffice. Finally, 9 percent of the unemployed in January had voluntarily left their last jobs. (See table A-6.)

Under the new survey procedures, the number of persons employed part time for economic reasons--sometimes referred to as the partially unemployed--was 5.2 million in January, substantially below the 6 to 6-1/2 million levels that had prevailed for about 2-1/2 years. The main reason for this large difference is that those so classified must now indicate explicitly their desire and availability for full-time work. (See table A-3.)

Total Employment and the Labor Force (Household Survey Data)

Total employment was 122.0 million in January, which is some 1.3 million higher than the figure for December. The bulk of this difference (approximately 950,000) is attributable to the introduction into the estimation process of 1990 census-based population controls, adjusted for the undercount. Some of the remaining difference may be associated with the introduction of the new survey questionnaire and collection methodology. The employment-population ratio--the proportion of the population with jobs--was 62.2 percent in January, only slightly higher than the figure for December. The January ratios were 55.7 percent for adult women, 72.4 percent for adult men, and 43.5 percent for teenagers. (See table A-1.)

Changes to the questionnaire now allow for the collection of data on multiple jobholders on a monthly basis. In January, 6.8 million persons, about 5.6 percent of all workers (on a not seasonally adjusted basis), held more than one job during the reference week. (See table A-8.)

Reflecting the net shift of persons from outside the labor force to both employment and unemployment under the redesigned survey, the labor force participation rate was 66.7 percent in January, somewhat higher than the proportion estimated using the former procedures. (See table A-1.)

Discouraged Workers (Household Survey Data)

The household survey redesign included a major change in the definition of discouraged workers. Two requirements were added: To be classified as discouraged, one must have searched for work during the prior year and be explicitly reported as currently available for work. Using this new definition, the number of discouraged workers was 600,000 in January, on a not seasonally adjusted basis. (See table A-8.) Under the former, less restrictive definition, the number of discouraged workers had held steady at about 1.1 million (seasonally adjusted) for over 2 years.

Industry Payroll Employment (Establishment Survey Data)

Nonfarm payroll employment edged up by 62,000 in January on a seasonally adjusted basis, as unusually frigid winter weather over much of the country during the reference week for the survey (the week of the 12th) limited the extent of job growth. During the prior 4 months, payroll employment gains had averaged 191,000. (See table B-1.)

Durable goods manufacturing employment rose for the fourth consecutive month, particularly in auto- and construction-related industries, including fabricated metals, lumber, furniture, and stone, clay, and glass products, as well as motor vehicle manufacturing itself. Durable goods employment increases have totaled 78,000 since September. Nondurables employment held steady, as rubber and plastics (also auto-related) and printing and publishing had sufficient gains to offset job losses in chemicals and petroleum products. Employment in the apparel industry was flat after extensive losses in recent months.

The number of construction workers was little changed, after seasonal adjustment; employment in the industry was depressed by the extreme winter weather in the Midwest and Northeast. Mining employment also was little changed, following large December gains attributable to the return of striking coal miners.

The trucking industry added 10,000 jobs to transportation, which has been experiencing strong growth in recent months. Wholesale trade also added 10,000 jobs over the month, mostly in durable goods distribution. While employment in retail trade increased by 20,000, there were losses-- probably weather-related--in general merchandise, apparel, and eating and drinking establishments. Jobs in automobile dealers and service stations rose by 11,000 in January, and their total has grown by 123,000 since January 1992.

The finance industry added 6,000 jobs over the month. Employment in the services industry was flat for the first time in nearly 2 years, as job gains of 25,000 in health services were overshadowed by weather-related losses in business services, amusements and recreation, and elsewhere in the industry. Employment of Federal workers declined by 20,000, as temporary postal workers hired for the holiday season were released.

Weekly Hours (Establishment Survey Data)

The average workweek for production or nonsupervisory workers on private nonfarm payrolls rose 0.3 hour to 34.8 hours in January, seasonally adjusted. The manufacturing workweek and overtime hours remained at high levels of 41.7 and 4.4 hours, respectively. (See table B-2.)

The index of aggregate weekly hours of private production or nonsupervisory workers on nonfarm payrolls rose by 0.9 percent to 126.3 (1982=100) in January, largely a result of the longer workweek. The manufacturing index increased by 0.3 percent to 102.4. (See table B-5.)

Hourly and Weekly Earnings (Establishment Survey Data)

Average hourly earnings of private production or nonsupervisory workers on nonfarm payrolls increased by 0.7 percent in January, after seasonal adjustment. Average weekly earnings increased by 1.6 percent. Before seasonal adjustment, average hourly earnings rose 10 cents to $11.07 and average weekly earnings were down 96 cents to $379.70 in January. Over the year, average hourly earnings increased by 2.8 percent and average weekly earnings by 3.7 percent. (See table B-3.)

Revisions to Household Survey Data

Redesign of the survey

Effective with the data for January 1994, estimates from the Current Population Survey (CPS) reflect the results of a major redesign of the survey. The redesign was undertaken to obtain more accurate and comprehensive information on the labor force activities of the population. The survey questionnaire has been revamped to include many new and revised questions regarding individuals' employment and unemployment activities, and an automated data collection environment has been introduced.

The new questionnaire and mode of data collection were tested using a separate sample of 12,000 households per month between July 1992 and December 1993. The parallel survey showed differences in the levels and ratios for many data series. For example, as shown in table B, the annual average unemployment rate for 1993 from the parallel test survey was 7.3 percent, compared with 6.8 percent in the CPS.

The higher overall rate realized in the parallel survey was due largely to an increase in measured unemployment among women, teenagers, and persons 65 years and over. Research to date suggests that the higher unemployment figures may be associated with the elimination of an implicit gender/age bias in the previous questionnaire. This bias may have led certain workers to underreport marginal job search activities. The new survey questionnaire is believed to elicit more accurate information on the labor market activities of all workers.

In addition to the changes in the questionnaire and collection methodology, the measurement of some concepts was changed as part of the redesign. The biggest change occurred in the definition of discouraged workers, which was altered to include the requirements that a person must have searched for work in the prior year and be currently available to take a job. The result is a greatly reduced number of discouraged workers. Similarly, the measurement of persons employed part time for economic reasons (those who would prefer full-time work) was tightened by adding two criteria for persons who usually work part time: They must be explicitly reported as desirous of and available for full-time work. Previously, such information was inferred. The result is a smaller number of involuntary part-time workers. Many other changes have been made to improve the accuracy of the survey and to provide additional information on labor force topics.

Introduction of 1990 census population controls

Also with the release of data for January 1994, population controls projected forward from 1990 decennial census counts and adjusted for the estimated population undercount are being introduced into the CPS. From January 1982 through December 1993, monthly intercensal population controls had been based on projections carried forward from the 1980 census. The introduction of new population controls raises the total population levels and particularly increases the proportion of the population that is Hispanic.

As shown in table C, when applied to 1993 annual average data, the new population figures raise the level of the civilian noninstitutional population by about 1.3 million, the civilian labor force by 1.1 million, employment by 950,000, and unemployment by 200,000, over published levels. In addition, the unemployment rate was raised by 0.1 percentage point, reflecting the overall increase in the Hispanic share of the population. (Hispanics have above-average unemployment rates.) While data for 1993 and earlier years have not yet been revised to reflect the introduction of these new population controls, revisions to historical data will be made at some future date.

The combined effects of the CPS redesign and the introduction of 1990 census-based population controls for selected 1993 annual averages are shown in table D. For a comprehensive examination of these effects, see "Revisions in the Current Population Survey Effective January 1994" in the February 1994 issue of Employment and Earnings.

Changes in data presentation

The redesign has necessitated changes in the publication of household survey data in this release, as some series have been affected by definitional changes, new series have been created, and others will no longer be published. Additional publication changes not directly related to the redesign of the CPS also have been made at this time. The most important changes are described below.

Discontinued tables

Old table A-7, Range of unemployment measures based on varying definitions of unemployment and the labor force (also referred to as alternative unemployment measures U-1 through U-7), has been temporarily discontinued. The new questionnaire results in many definitional and measurement changes for several of these series. Additionally, seasonally adjusted data for discouraged workers are currently not available. BLS will conduct research into alternative unemployment measures and reintroduce a revised range of alternative measures in the near future.

Old table A-9, Employment status of male Vietnam-era veterans and nonveterans by age. These data have been discontinued in this release but will continue to be published in Employment and Earnings.

Old table A-11, Persons not in the labor force by reason, sex, and race (published quarterly) has been discontinued. Newly available monthly data on persons not in the labor force, presented in table A-8 of this release, have replaced the quarterly estimates formerly contained in table A-11.