U.S. Federal Labor Relations Authority

Search form


United States of America




In the Matter of







Case No. 95 FSIP 150


    Local 2077, American Federation of Government Employees, AFL-CIO (Union), and the Department of the Air Force, Selfridge Air National Guard (ANG) Base, Selfridge ANG Base, Michigan (Employer) filed a joint request for assistance with the Federal Service Impasses Panel (Panel) to consider a negotiation impasse pursuant to the Federal Employees Flexible and Compressed Work Schedules Act of 1982 (Act), 5 U.S.C. § 6120 et seq., to resolve an impasse arising from the Employer’s determination to terminate a compressed work schedule (CWS).

    Following an investigation of the request for assistance, the Panel determined to assert jurisdiction over the case and to assist the parties through an informal conference with a Panel representative. The parties were advised that if no settlement were reached, the representative would report to the Panel on the status of the dispute. After considering the report and the recommendations of its representative, the Panel would take final action in accordance with 5 U.S.C. § 6131 and 5 C.F.R. § 2472.12 of its regulations.

    Pursuant to the Panel's determination, Panel Member Edward F. Hartfield met with the parties at Selfridge ANG Base, Michigan, on December 4, 1995. During the conference, Member Hartfield toured the Precision Measurement Equipment Laboratory (PMEL), and interviewed two bargaining-unit employees and two supervisors who work in the laboratory. The parties, however, did not resolve the dispute during the course of this procedure. The Employer provided a data package to Member Hartfield on the day of the conference. In addition, the Employer also introduced several letters from other military facilities that described their experience with flexible work schedules. In accordance with Mr. Hartfield’s directions, the Union submitted a written statement on December 6, 1995, to address those documents the Employer provided near the end of the informal conference. Member Hartfield has reported to the Panel and it has now considered the entire record, including written materials submitted by both parties.


    The Employer maintains a mission-capable, combat-ready force of F-16 fighter aircraft. Selfridge ANG Base is one of three Air Force bases in the United States owned and operated by the National Guard, but it is under the Department of the Air Force. The Union represents 570 bargaining-unit employees at pay grades GS-2 through -11 who work in support positions providing aircraft maintenance, snow removal, and other services needed for operating an Air Force Base. The 19 employees in the PMEL are highly-trained technicians who are engaged in calibrating, repairing, aligning, and recalibrating navigational instruments and tools such as torque wrenches for 21 military organizations in 7 states. At 180-day intervals, they also calibrate jet engine test cells at a number of airfields from a mobile laboratory housed in a trailer. Laboratory employees work a variety of schedules: the 4 employees who will be affected by the outcome of this dispute are on a 4-10 CWS; 1 employee is on a 5-4/9 CWS; and 14 employees are on 8-hour days.(1) The parties’ collective-bargaining agreement (CBA) is due to expire on March 13, 1998.


    The dispute concerns the Employer’s determination to terminate the 4-10 CWS in the PMEL. The specific issue before the Panel is the following:

Whether the finding on which the Employer has based its determination to terminate the 4-10 CWS is supported by evidence that the schedule has caused an adverse agency impact as defined under the Act.(2)


1. The Employer's Position

    Essentially, the Employer asserts that the Panel should find that the evidence on which it bases its determination to terminate the 4-10 CWS for four employees in the PMEL establishes that the schedule has caused an adverse agency impact as defined under the Act. Performance data compiled at two intervals over the last year indicate that the productivity of the 4 employees on the 4-10 CWS is significantly lower than that of the 14 employees on a regular 8-hour, 5-day-a-week schedule. In this regard, for the 11 months ending on June 26, 1995, employees on the 10-hour shifts, on average, completed 43.02 units of work per month as compared with 55.01 units completed by employees on 8-hour shifts.(3) The aggregate data reveal that employees on 10-hour shifts accomplish 22 percent less work than employees on 8-hour shifts. Since each group receives a share of work that "represents a cross-section of equipment by complexity and time requirements, i.e., from tire gauges and micrometers to complex integrated electronics test sets," the comparison of work unit completion totals is not skewed by differences in the difficulty and complexity of the work each employee actually performed. In addition, the backlog levels have steadily risen from 8 days in January 1995 to 16 days in November 1995; based on current trends, if changes are not made to employees’ work schedules, the backlog is projected to reach 18 days within the next 6 to 8 months. Furthermore, the productivity of employees on 10-hour shifts is also diminished because they are without their immediate supervisor for 2 hours each day. Regarding the use of sick leave, employees on the 10-hour shift use an average of 23 percent more sick leave than employees on the 8-hour shift who begin work at 0700 or 0800.(4) Essentially, the data establish that the existing 4-10 CWS has caused a reduction in productivity, and that its termination would improve timely delivery of customer services and reverse the trend toward a growing backlog.

2. The Union's Position

    The Union would have the Panel find that the 4-10 CWS for four employees in the PMEL has not had an adverse agency impact. The 4-10 CWS is desirable because two employees have commuting distances of 150 and 400 miles, respectively. While they stay on the base during the week, the CWS greatly improves their family life on the weekends. The Employer’s productivity data are flawed because they do not take into account the practice by some employees of taking much of the easy work, called "ducks," off the shelves, leaving the more complex, time-consuming calibrations for others. Without greater precision regarding the ratio of complex to simple assignments per employee, the Employer’s figures compare apples to oranges. Regarding the increase in the work backlog, it is directly related to the deleterious effect of the June 1995 decision to rotate mechanical and electrical engineering work assignments (see footnote 3) and to avoid overtime use beginning in October 1995; now, employees not only do their own work, but also must consult with co-workers who have taken on their former assignments. In addition, the loss of four laboratory employees who were riffed in January and March, 1995, is not balanced by the elimination of one customer, the Pittsburgh AFB. Trends since 1992 show that loss of personnel is always associated with an increase in the backlog. Finally, the Employer improperly blames the schedule for productivity lags during periods when an immediate supervisor for one side of the laboratory is not present; instead such problems should be attributed to ineffective supervision.


    Under § 6131(c)(2) of the Act, the Panel is required to take final action in favor of the determination of the agency head or, in this instance, that of his delegatee, to terminate the CWS if the finding on which it is based is supported by evidence that the schedule has caused an "adverse agency impact." The Act's legislative history clearly indicates that the Employer bears the burden of proving adverse impact.(5)

    Having considered the record before us, we find that the Employer has not met its statutory burden under the Act. Preliminarily, we acknowledge that the Employer made a serious effort to assemble relevant data from its performance management system to support its finding that a significant decrease in productivity and increase in the backlog of work are related to the 4-10 CWS. In our view, however, the following deficiency raises credible doubts about the reliability of the data. In this regard, the Employer does not explain the methodology it applied in comparing the productivity figures for employees on different schedules. Thus, it has failed to support persuasively the statement in its chief supporting document that the "items of work are distributed in complexity so that each group, 10-hour shift or 8-hour shift, represents a cross section of equipment by complexity and time requirements." Rather, our examination of the 12 sanitized records of individual employees’ output that the Employer provides lends support to the Union’s position that the work employees perform may be distributed unevenly.(6) Moreover, while the Employer concludes that the rising backlog is attributable to the 4-10 CWS, the evidence suggests that it is as likely to have occurred because of the June 1995 rotation of employees’ work assignments and the Employer’s October 1995 decision to avoid assigning overtime. Finally, both parties admit that there are problems with employees’ lack of productivity during the first and last hours of the day when an immediate supervisor is not present. We are convinced that management has within its discretion alternative ways of curing such problems other than the termination of the 4-10 CWS schedule. In summary, we conclude that the evidence provided does not support the Employer’s finding that the 4-10 CWS is causing an adverse agency impact, and shall order the Employer to rescind its determination to terminate the 4-10 CWS in the PMEL.


    Pursuant to the authority vested in it by the Federal Employees Flexible and Compressed Work Schedules Act, 5 U.S.C. § 6131(c), the Federal Service Impasses Panel, under § 2472.12(b) of its regulations, hereby orders the Employer to rescind its determination to terminate the 4-10 CWS.


By direction of the Panel.

Linda A. Lafferty

Executive Director

December 15, 1995

Washington, D.C.


1.Under the 4-10 CWS, employees work four 10-hour days and have 1 permanently-assigned day off each week and under the 5-4/9 CWS, employees work eight 9-hour days, one 8-hour day, and have 1 day off every other week.

2.Under § 6131(b), "adverse agency impact" is defined as:

(1) a reduction of the productivity of the agency;

(2) a diminished level of the services furnished to the public by the agency; or

(3) an increase in the cost of agency operations (other than a reasonable administrative cost relating to the process of establishing a flexible or compressed work schedule).

The burden of demonstrating that implementation of a proposed CWS is likely to cause an adverse agency impact or, as in this case, that an existing CWS has had an adverse agency impact, falls on an employer under the Act. In the instant case, the head of the agency, the Secretary of Defense, delegated the needed authority to make the requisite finding under the Act through the Secretary of the Air Force to the Commander of the installation. Air Force Instruction 36-102, February 18, 1994, and AFI 36-807/AFI Supplement, June 15, 1995, convey the delegation of authority.

3.In June 1995, the types of assignments given to employees were rotated; employees in the mechanical section were given electrical engineering tasks and those in the electrical section were given mechanical engineering tasks. Many employees had minimal recent experience in the new areas. For the period ending on November 8, 1995, the productivity of employees on 10-hour shifts, on average, fell by 7.77 items completed to 35.25; the productivity of employees on 8-hour shifts dropped, on average, by 4.72 items to 50.29. The aggregate revised data for employees on 10-hour shifts indicate that they are 30 percent less productive than employees on 8-hour shifts.

4.The Employer indicates that employees on 8-hour shifts who begin work at 0600 hours also use a larger percentage of sick leave.

5.See 128 CONG. REC. H3999 (daily ed. July 12, 1982)(statement of Rep. Ferraro); and 128 CONG. REC. S7641 (daily ed. June 30, 1982)(statement of Sen. Stevens).

6.These raw data list the number of completions in equipment categories J, F, and K: category J represents equipment involving codes which is the least difficult and least time consuming to work on; category F represents more complex operations involving functions; and category K represents equipment where trouble shooting is required.