INTRODUCTION

Why this survey was made

On September 21, 1970, Senator William Proxmire wrote the Comptroller General urging that a comprehensive evaluation be made of the possibilities for measuring productivity in the Federal sector of the economy. In his letter, Senator Proxmire made reference to a study by the Bureau of the Budget in the early 1960's. This study, which analyzed the potential for constructing overall productivity indices in five major governmental agencies, concluded:

"As a result of the study, it is believed that
development of valid productivity measures is
feasible for a considerable proportion of Fed-
eral Government activities. The principal ob-
stacle in Government is the requirement that
products or services be measurable over a pe-
riod of time on a consistent basis." [This is
possible] "*** if some form of cost accounting
or work measurement system is in effect."1

Preliminary inquiry revealed (1) that this promising beginning was not continued because the staff assigned was transferred to other functions in 1965 and (2) that there was no program currently being conducted by the executive branch to develop overall measures of productivity. Consequently the Comptroller General invited the Director of the Office of Management and Budget and the Chairman of the Civil Service Commission to join with the GAO in a project aimed not only at responding to Senator Proxmire's request for "an evaluation of the possibilities" but, if possible, to demonstrating the feasibility of achieving this objective and to searching out actions to enhance productivity in the Federal sector of the economy. These officials immediately accepted the Comptroller General's invitation.2

1

Measuring Productivity of Federal Government Organizations, 1964, page 4.

2 See appendix A.

THREE-PHASED APPROACH

A steering committee of the three participating agencies was formed in March 1971, consisting of Executive Director Bernard Rosen of the Civil Service Commission, Assistant Director Dwight Ink of the Office of Management and Budget, and Assistant to the Comptroller General Tom Morris. This Committee appointed a joint staff to carry out the project, and invited 17 agencies1 of the executive branch to join in a three-phased cooperative effort as described in Exhibit 1-1 below.

In early April 1971, Phase I was launched at a planning meeting with the 17 agencies, at which time each was requested to submit an inventory of its current usage of quantitative measurement systems, showing the percent of its civilian employees whose work effort is reasonably covered by such systems.

In order to assure that all significant measures were identified in the inventory, the agencies were asked to cover the following four types.

1The Cabinet Departments, Postal Service, NASA, AEC, GSA, VA, and SEC. NLRB contributed to the measurement project described in Chapter 2.

2After Phase I was completed effectiveness measures were added to the scope of the project.

1. "Overall Productivity Indexes" --These are measures
of the final physical outputs of an entire organiza-
tion (or component, subcomponent, etc.) divided by
the physical inputs, in order to produce a produc-
tivity index which can be consistently computed from
year to year in real terms (i.e., in constant dol-
lars). Labor inputs will be used in all cases.
Capital and other cost inputs will be included where
significant to the activity (as in the case of in-
dustrial facilities and costly data processing in-
stallations).

2. "Work Measures" --These are measures of the physical work units produced by individual work centers (within an organization or component) which can be compared to an objectively derived standard of performance to assess performance efficiency on a current basis (such as daily, weekly, monthly). Example: Number of pieces of first class mail processed per hour in a Post Office, compared to an established standard.

3. "Unit Cost Measures"--These are obtained by relating physical work units produced to corresponding costs. 4. "Manpower Planning Measures"--These are systems of forecasting manpower requirements based on statistical data in respect to workload, activity rates, peakload requirements, safety standards, etc. (An example is FAA's technique of staffing Air Traffic Control towers, based on peak-hour traffic and human stress factors.)

All agencies invited agreed to participate in Phase I, and by late May the requested inventories had been submitted. Briefings were then held with agency representatives to discuss their practices in developing and using each type of

measure.

Growing out of the Phase I review was a presentation to the principals in June 1971, by the steering committee, and a memorandum report (see app. B). The committee recommended to the Director of OMB, the Chairman of the Civil Service Commission, and the Comptroller General that fiscal year 1972 be devoted to a continued cooperative effort with the 17 agencies aimed at

--first, testing the feasibility (with assistance

from the Bureau of Labor Statistics) of constructing overall productivity indices,

--second, conducting case studies and workshops to re-
veal opportunities for improving the use of existing
measurement data, and

--third, making other special studies of ways to en-
hance productivity in the Federal sector.

The proposals were endorsed by the principals, and Phase II was launched at a conference of the participating agencies on September 8, 1971.

The accompanying report sets forth the findings and recommendations which have resulted from the Phase II stud

ies. As background to the report, the remainder of this introduction briefly describes:

--The key findings from the inventory conducted in
Phase I.

--The principal research projects conducted in Phase II.

KEY FINDINGS OF THE PHASE I INVENTORY1

Exhibit 1-2 below shows a statistical summary of the usage of various types of measurement systems, as disclosed by the participating agencies.

1See appendix D for a brief history of Federal Productivity

measurement.