Skip Navigation

Mr. Chairman: I appreciate this opportunity to be with you today to discuss the use of productivity data in evaluating federal programs.

Federal productivity data measure the output--in terms of goods and services--per employee in certain activities. An example that comes to mind is the annual number of passports an employee issues. Such information is collected for the activities of about two out of every three federal civilian federal employees. (This includes some 684,000 employees of the U.S. Postal Service, which is an off-budget agency. The coverage for nonpostal civilians is 55 percent.)

The Bureau of Labor Statistics (BLS) is responsible for the government-wide productivity reporting system. Federal agencies annually submit raw data to the bureau for productivity calculations. The data are aggregated into 28 measurement categories such as "Social Services and Benefits," "Records Management," and "Printing" (The categories used do not conform to those of the budget.) Summary information on productivity in these categories is submitted to the Congress in the annual report of the National Center for Productivity and Quality of Working Life.

Another major source of productivity data is the budgetary justifications of individual agencies. In accordance with Circular A-ll from the Office of Management and Budget, agencies are required to base their budgetary estimates on detailed analyses of manpower requirements. The analyses are supposed to take into account the effects of capital investment proposals, as well as workload requirements, productivity trends, and changes in the quality of output. As Mr. Staats indicated, there are wide variations among the ways individual agencies fulfill these requirements for budgetary submissions.

This Task Force is primarily interested in the possible use of productivity data in the Congressional budget process. In this connection, I would like to touch on three subjects:

• Productivity information as an evaluation tool;
  
  
• Incentives for using productivity data; and
  
  
• Opportunities for the Congress to make further use of productivity data.

When the work of an agency is fairly routine, and a quantitative account of employee performance accurately reflects the agency's work, then productivity data can be a useful tool for evaluation. In general, the operations of the Postal Service are especially well suited to this evaluative technique. Clerical support, automatic data processing, and finance and accounting lend themselves well to such assessments. Roughly two-thirds of the federal civilian workforce is engaged in such activities. There are cases, however,--such as policy analysis, research, and evaluation--in which employee performance cannot be usefully quantified. In these instances, measuring productivity is inappropriate and possibly even counterproductive.

In both the public and the private sectors, sensible use of productivity measurement requires an awareness of its limitations. It is difficult to get productivity measures to reflect changes in the quality and composition of programmatic outputs. If changes require much greater investments of time, misleading productivity measures can result. For example, the workload of the Federal Bureau of Investigation and U.S. attorneys has shifted from cases such as car theft to more complex areas such as drug enforcement, monopoly and antitrust actions, and white collar offenses. These latter kinds of cases demand a great deal of research and court time. Productivity, as measured by BLS, is limited to just one resource: labor. But both labor and capital investment in facilities and equipment are resources used in government and private sector programs. When the cost of capital investment is significant, mere measures of labor investments can be misleading. The Social Security Administration (SSA) can serve as an example. The SSA has a substantial capital investment in automated data processing equipment, which has brought about marked increases in the number of claims an SSA employee processes. Thus, in the case of the SSA, labor productivity alone cannot reflect the administration's true efficiency; capital costs also must be considered.

In both the public and private sectors, productivity measurement is of little value unless it is used in decisionmaking. In the federal government, there is a lack of incentive to use the data how being collected. A federal program manager often views productivity measurement as a no-win proposition. If he reports losses in productivity, his management abilities may be questioned. If gains are reflected, his resources may be transferred elsewhere.

Nor do federal employees benefit from programs to improve productivity. Studies in the private sector have found that productivity tends to go up when pay is tied to performance. Present methods of setting federal pay rates do not reinforce productivity increases with monetary incentives. The annual comparability pay adjustments that federal employees get do not reflect employee performance--good or bad. Likewise, the within-grade pay raises (the Civil Service Commission's "step increases") are virtually automatic inasmuch as they are tied to longevity rather than merit. Historically, only 2 percent of all federal employees have been denied these longevity increases because their work was found to be below minimum standards. Also, although financial incentives for employee suggestions and outstanding performance are available, they are seldom used. In fiscal year 1976, for example, cash awards for employee suggestions and superior performance represented 0.06 percent of total payroll.

The establishment of a zero-based budget review in the Executive Branch offers an excellent opportunity for making selective use of productivity data in evaluating programs. The real pay-off is not so much in measuring the increase or decrease in the rate of productivity, but rather in understanding why such changes have occurred. Thus, if productivity assumptions and performance were identified and related to the President's budgetary request, more meaningful long-term trend analyses could be conducted. These analyses could help signal areas where problems exist or where greater efficiency has been achieved. Significant productivity changes could receive more extensive analysis concerning technology, capital investment, and qualitative aspects of the work processes. Productivity measurement can be particularly useful if it is accompanied by an analytical capability that looks behind the numbers.

Productivity data can be a useful tool for the Congress in evaluating agency programs. The Authorizing and Appropriations Committees are probably in the best position to make use of productivity measurements. At this level, programs can be examined in detail, including the efficiency with which resources are utilized.

I would not expect productivity data to be very helpful to the Budget Committees themselves. Their concern is with the allocation of resources among broad functional areas, in which the usefulness of productivity information would be lost. Moreover, the potential budgetary impact is very small.

Contrary to public belief, the number of federal civilian workers (not counting postal employees) has dropped slightly during the last 10 years. Full-time, permanent civilians in the Executive Branch numbered 1.9 million at the end of fiscal year 1977, compared to 2.1 million in 1968. About half of these present employees are covered by the productivity reporting system. Their pay represents about 5 percent of total budgetary outlays. The remaining 95 percent goes for benefit payments to individuals, grants to state and local governments, the armed services, other nonlabor costs, and civilian personnel for which productivity is not measured. Thus, the potential budgetary impact of using productivity measurements is relatively small compared to total outlays.

The Budget Committees, however, could draw attention to important areas in which the Congress can act to increase the prospects for improving productivity. I would like to outline four areas in which there are significant opportunities:

• Employee Compensation. The recent Rockefeller Commission on Federal Compensation recommended that within-grade pay increases be tied to merit rather than longevity. The Administration is considering proposals in this area--at least for top officials and middle-grade managers. This is a possible area for Congressional review and legislative action. Changes in the way step increases are awarded (to whom, and with what increases in compensation) could have far-reaching implications for greater productivity.
  
  
• Flexitime. Flexible work time was first introduced at the Bureau of Indian Affairs in 1972 and at the Social Security Administration headquarters in April 1974. Today, 12,000 of all headquarters Social Security employees (about half) are covered. Flexitime was introduced at Social Security because of a high incidence of short-term absences: heavy use of sick leave, leave without pay, and a dependence by the employees on local transportation. The high leave rates were assumed to be related to child care, since about 80 percent of the work force are women. Signs of increased output with flexitime occurred almost immediately: an evaluation of the program in 1974 indicated an annual increase in productivity of 11.6 percent. Other agencies have been testing the productivity implications of flexitime, Experiments with four-day, forty-hour work schedules have been recommended by the General Accounting Office. (Such tests would require legislation to avoid costs relating to overtime pay.)
  
  
• Reorganization. Proposals for organizational change are an important part of the Administration's program. Excessive overhead, unnecessary levels of review, and limited autonomy of managers are often factors that impede productivity. Reorganization is one means of addressing these problems. The Congress can give close attention to the productivity implications of the Administration's proposals.
  
  
• Sunset Legislation. The review provisions of S-2 offer the Congress another opportunity for evaluating resource utilization in specific programs. The basic question with respect to many federal activities is not whether they are needed, but rather how best they can be carried out. Productivity measurement can be a useful tool in addressing this question. The evolutionary approach anticipated for implementing the sunset concept would allow flexible Congressional application of productivity data.

In conclusion, Mr. Chairman, I would like to reiterate my support for greater emphasis on productivity data in the Legislative as well as the Executive Branch. The Budget Committees can encourage greater use of existing data and can help raise the level of expectation for use of productivity data by federal agencies. Such attention could enhance the quality of decisionmaking and help stretch federal dollars to help meet competing demands on the budget.

Thank you, Mr. Chairman.