Printer Friendly

The Employee Turnover and Job Openings Survey.

Insufficient numbers of workers in some occupations have long been recognized as potential bottlenecks in the economy. As a partial solution, a longstanding policy gives immigration preference to workers with specific skills that are difficult to obtain in the domestic labor market. The process of identifying such workers is cumbersome, however, because certifications are done on a sace-to-case basis. In order to improve the process, the Immigration Act of 1990 authorized the Department of Labor to test a program that would "precertify," on the basis of labor market information, occupations with shortages of workers. The Senate Committee on Appropriations, in the appropriation for the Department of Labor for fiscal year 1990 (H.R. 2990), also directed the Department to earmark funds to "develope a methodology to annually identify national labor shortages."

The Employment and Training Administration, in response to these mandates, commissioned the Employee Turnover and Job Openings Survey to evaluate the possibility of collecting data on job opening, turnover, and marginal wages. (1) In the opinion of many labor market analysts, such demand-size data are important components for determining which occupations have labor shortages.

The issue these data could address go beyond labor shortage concerns. For example, many economists believe that structural change is an important determinant of the unemployment rate. That is, the unemployment rate has gradually risen over time, in part because a growing mismatch exists between the skills of unemployed workers and the skill requirements of vacant jobs. The mismatch is due principally to an increase in the rate of change in the industrial composition of employment. Such trends can be better quantified by the analysis of job openings and turnover data. In the same way that national vacancy data could help economists understand the source of fluctuations in the aggregate unemployment rate, information on job openings by region or State would be helpful for identifying the sources of regional variations in unemployment and for understanding patterns of migration.

Aware of their potential utility, public policymakers and research economists have often cited the need for the government to collect statistics on job openings, along with the information on turnover of employees and duration of vacancies needed to make meaningful analyses of those statistics. In 1962, the President's Committee to Appraise Employment and Unemployment Statistics noted that one of the most frequently mentioned suggestions for improving the body of labor market knowledge was instituting a program on job vacancy statistics. (2) In 1979, the National Commission on Employment and Unemployment Statistics recognized the conceptual appeal of data on job vacancies, but, purely on considerations of cost, recommended against establishing such a program. (3) More recently, in 1989, Profesor Sar Levitan, who chaired the National Commission, wrote in a report to the Joint Economic Committee, "An ongoing survey of job openings could shed light on the availability of jobs for the structurally unemployed and provide a timely warning of economic downturns." (4) Levitan then, however, reiterated that the collection of statistically reliable vacancy data would entail considerable difficulty and expense.

Survey purpose and design

Conducted as a pilot project from late 1990 to mid-1991, the Employee Turnover and Job Openings Survey sought to determine whether advanced data collection technologies and a specific legislative mandate to produce data related to national labor shortages could lead to a cost-effective statistical program. (5) The data obtained from the survey would, or course, be of interest, but the main purposes of the project were to assess the data concepts and methods and to estimate the costs of implementing a full-scale survey on the national level.

The survey provided the data to develop two direct measures of the difficulty employers found in hiring prospective employees: the duration of existing job openings and the "vacancy fill rate" (the number of new hires in a month divided by the number of jobs open at the end of that month). Economic theory also suggests that shortages in competitive markets will be accompanied, at the margin, by rising wages. Therefore, data were collected on the wages of new hires--the wages most directly affected by market conditions. Data on separations were collected to allow the information on job openings to be understood more completely in relation to turnover in an occupation.

The operational definitions of job openings and turnover used in the survey were designed as measures of specific activities. Job openings were vacant jobs for which there had been active recruiting and in which work could be started immediately. The definitions of new hires and separations excluded short-term events, such as temporary layoffs or absences, focusing instead on permanent separations and new hires. New-hire wages were defined as the arithmetic mean wage of persons hired during the month. The complete definitions of job openings, separations, new hires, and new-hire wages used in the survey are presented in the box on this page.

The survey questionnaire was constructed using the latest literature on document design. The final instrument reflected input from focus groups drawn from the business community and the results of a field pretest. The pilot survey covered only eight industries, one from each major nonfarm industry group. (6) Each industry-specific questionnaire required a customized occupational matrix, which included all occupations accounting for at least 0.25 percent of total industry employment.

Data collection

Data collection activities began on October 1, 1990, with an operations test--a preliminary test to see how the survey would be carried out in practice--on 100 establishments. After incorporating systems and procedural improvements suggested by the results of the operations test into the data collection strategy, survey staff in the national office commenced full-sample operations in December 1990. (7) Data were collected via the mail, with

follow-up by mail and computer-assisted telephone interviews. Telephone interviews were also used to assure the internal consistency of questionnaires received by mail.

The data were collected in three rotations of approximately 1,000 establishments each. The rotation plan resulted in a unit being in sample for 1 month, out of sample for 2 months, and in sample again for 1 month. The months for which data were collected were November 1990 through April 1991. November, December and January made up the first rotation through the entire sample, and February, March, and April made up the second. Response rates were higher in the second rotation, as interviewers became more experienced and the name-and-address file became more refined. The second rotation yielded a usable response rate of 75 percent, versus 70 percent in the first.

Results and analysis

The central finding of the survey was that the collection of data on job openings and employee turnover remains a


difficult and labor-intensive understaking. In general, this confirms the findings of earlier tests, including those of the Job Openings Pilot Program issued in 1981. (8) Based on the results of the survey, a full-scale program to produce, on a national scale, the demand-side data required for annual analyses of occupations experiencing a labor shortage could be conducted by the BLS for about $11-$12 million a year.

The responses to the survey were evaluated by a response analysis survey, conducted from April through June 1991. The response analysis survery found that a substantial majority of respondents used personnel or payroll records as the primary sources of information to complete the Employee Turnover and Job Openings Survey. As a result, there were no major problems with the validity of the data, whether collected by mair or by computer-assisted telephone interview, although technical issues of multiple reference periods, the treatment of


interestablishment transfers, and the calculation of wages under nonstandard pay schemes would need to be addressed during the implementation of a full-scale program.

Although the primary objectives of the Employee Turnover and Job Openings Survey were to assess the technical feasibility and estimate the cost of conducting a full-scale program, it yielded statistical results for analysis as well. The following overview of the findings pertain strictly to the aggregation of the eight specific industries that were selected for the survey and do not reflect estimates for the entire economy.

In the eight industries surveyed, there were numerous openings in the first rotation (November 1990-January 1991) in professional and technical, service, and production and related jobs. (See table 1.) In the second rotation (February 1991-April 1991), there was a statistically significant decrease in job openings among the production and related jobs. This shift may reflect seasonal factors, irregular events, or cyclical developments. The survey was conducted in the midst of a recession that resulted in a sharp reduction in employment in several of the industries included in the sample. These declines may have affected the number of openings, particularly for production jobs, which tend to be cyclically sensitive.

A large number of job openings does not, in itself, signal a shortage of labor. More important are the length of time such openings remain unfilled and the number of openings relative to new hires (the fill rate). When these criteria are applied, it becomes clear that the professional and technical occupations and the managerail occupations most likely had a shortfall of labor. For each of these two groups, more than half of the job openings had been open for more than 4 weeks in both rotations. By contrast, little more than one-tenth of the job openings for the service occupations were of long duration. In addition, the professional and technical occupations and the managerial occupations had relatively little new hiring and separations. The fill rates were less than 1 for both groups, indicating that openings exceeded hirings.

The large number of job openings in the service occupations, on the other hand, were also associated with high levels of separations and new hiring. Service occupations made up about two-fifths of the separations and nearly half of the hiring in both rotations.

Future prospects

As noted before, the pilot program was too limited in industrial scope and the sample was too small to carry out a definitive analysis of occupational labor shortages. Rather, it was intended solely to highlight some of the tools that would be available for such an analysis from a full-scale survey program. With data from a fully operational survey, research could:

1. Determine the relative importance to be accorded the various statistics on turnover and job openings in the identification of labor shortages;

2. Develop a long enough series of data to gauge the trends in important indicators, such as the change in earnings of new hires;

3. Compare the lists of occupations with a shortage of labor developed with the data from the Employee Turnover and Job Openings Survey with lists developed using other methodologies.

The full report, Employee Turnover and Job Openings Survey: Results of a Pilot Study on the Feasibility of Collecting Measures of Imbalances of Supply and Demand for Labor in an Establishment Survey, is available from the Office of Employment and Unemployment Statistics, Bureau of Labor Statistics, Washington, DC 20012.


(1) Bureau of Labor Statistics, Employee Turnover and Job Openings Survey: Results of a Pilot Study of the Feasibility of Collecting Measures of Imbalance of Supply and Demand for in an Establishment Survey (Washington, Bureau of Labor Statistics, 1991). See Chapter III for a discussion of a theoretical foundation for using such data in analyzing labor shortages.

(2) President's Committee to Appraise Employment and Unemployment Statistics, Measuring Employment and Unemployment (Washington, Government Printing Office, 1962).

(3) National Commission on Employment and Unemployment Statistics, Counting the Labor Force (Washington, Government Printing Office, 1979), p. 122. This report is summarized in Robert L. Stein, "National Commission recommends changes in labor force statistics," Monthly Labor Review, April 180, pp. 11-21. See also Harry Frumerman, "Job Vacancy Statistics," in National Commission on Employment and Unemployment Statistics, Concepts and Data Needs--Appendix Volume I (Washington, Government Printing Office, 1979), pp. 602-35.

(4) Sar A. Levitan and Frank Gallo, Workforce Statistics: Do We Know What We Think We Know--and What Should We Know? (Washington, Joint Economic Committee, 1989).

(5) Earlier efforts, such as the Job Openings and Labor Turnover Survey of the late 1960's and the Job Openings Pilot Project of the late 1970's, were hampered by very broad goals and were conducted without the availability of advanced, computer-assisted survey techniques.

(6) The eight industries (and their SIC numbers) were oil and gas extraction (13), special trade contractors (17), electronic and other electrical equipment (36), trucking and warehousing (42), machinery, equipment, and supplies wholesaling (508), eating and drinking places (58), depository institutions (60), and hospitals (806).

(7) The States of Maine and Georgia also participated in the project. The experience of the Maine State Employment Security Agency is summarized in Maine Department of Labor, Employee Turnover and Job Openings Survey: An Analysis of the Methodology (Augusta, ME, Division of Economic Analysis and Research, 1991).

(8) Bureau of Labor Statistics, Job Openings Pilot Program: Final Report (Washington, Bureau of Labor Statistics, 1981).

Richard M. Devens, Jr., is a senior economist in the Office of Employment and Unemployment Statistics, Bureau of Labor Statistics.
COPYRIGHT 1992 U.S. Bureau of Labor Statistics
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1992 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Devens, Richard M., Jr.
Publication:Monthly Labor Review
Date:Mar 1, 1992
Previous Article:Statistical needs in Eastern Europe.
Next Article:Benefits in state and local governments address family concerns.

Related Articles
New tools for labor market analysis: JOLTS: as a single, direct source for data on job openings, hires, and separations, the Job Openings and Labor...
JOLTS starts. (Labor Month in Review).
The March review.
Obtaining information from the Bureau of Labor Statistics.
Job openings and labor turnover.
Labor turnover.
Labor turnover.
Labor turnover and employment in different U.S. regions.

Terms of use | Copyright © 2017 Farlex, Inc. | Feedback | For webmasters