Streamlining Management Control

Recommendations and Actions

SMC01: Implement a Systems Design Approach to Management Control


A striking characteristic of the federal government is its constant struggle to control a conglomerate of agencies several times larger than any existing holding corporation. It uses a labyrinth of disjointed mechanisms--put in place by both the executive branch and Congress--in an attempt to achieve control. The existing control mechanisms were developed at different points in time, primarily as individual reactions to specific problems. In aggregate, they were patterned after corporate control structures of the Industrial Era in the 1930s--large, centralized, top-down. These premises are no longer appropriate in an Information Age environment of rapid change and technology that makes information readily accessible to managers and employees at many different levels in an organization. The existing system provides the illusion of control. But in fact, these mechanisms overlap, duplicate, confuse, and sometimes contradict each other. Even a cursory look at the federal approach to management reveals an overly layered and poorly coordinated set of controls characterized by ambiguity and often focused on unimportant things. And its cost is enormous--one estimate places it conservatively at $35 billion a year. As a result, the existing system steals precious resources from the real job of government: serving the customer.

Line Versus Staff Controls.

Control in the federal government is exerted by two groups--line management and staff offices. Line management, or "line control," is exercised by each agency's program offices, such as the managers in a local Social Security office. Typically, line managers monitor performance and appraise their employees, and they write procedures and ensure they are complied with. They also conduct formal program evaluations, and they develop internal controls. Oversight of line managers is performed by management levels above the program office within the agency, such as a regional Social Security office.

The sum total of line management control and oversight is very large and, many believe, overdone. There are over 278,000 supervisors and managers in the federal government--about 1 in 7 federal employees. This has contributed to horror stories of over-management; for example, the lengthy "chop chains" for sending an official letter, entailing as many as 10 or 15 formal check points, any one of which can change the contents. The result: A simple letter may take months to get sent.

Staff offices, on the other hand, can be both internal or external to an organization. "Internal" staff offices exert control in at least two ways--granting permission to act or second-guessing decisions. The first, granting permission to act, would include actions such as the general counsel staff saying that a proposed action is legal; a procurement office awarding a contract; or a personnel office classifying a job at a certain grade level. Staff control can also be exerted through second-guessing, or "oversight," which usually is a formal assessment of compliance with rules. Most internal staff oversight is conducted by program evaluation offices, the procurement office, the personnel office, and financial managers. Oversight methods include investigations, audits, reviews, evaluations, and inspections.

A variety of "external" organizations also have a direct or indirect role in staff oversight. These include Congress, the General Accounting Office (GAO), the Office of Personnel Management (OPM), the Office of Management and Budget (OMB), the General Services Administration (GSA), the Equal Employment Opportunity Commission (EEOC), the Environmental Protection Agency (EPA), the Merit Systems Protection Board (MSPB), the Office of Government Ethics (OGE), and the Offices of Inspectors General (OIG's). Currently, over 36,000 federal employees--not including Congress--conduct "external" oversight (see Appendix B for an illustrative listing of most of these staff organizations).

Control functions grew faster than other functions.

Over the past decade, control functions grew faster than the executive branch as a whole. While executive branch personnel grew seven percent between 1980 and 1990-91, employment in the personnel function increased 11 percent. In financial management, the increase was 27 percent and, in procurement, it jumped 60 percent. (See figure 1.)

Need for Change

Despite the fact that billions are spent on control, the public still considers waste, fraud, and abuse to run rampant in the federal government. The average American believes the government wastes 48 percent of every tax dollar.(1) Ironically, a contributing factor is the existing system of fragmented controls. A lack of accountability and the inappropriate focus of existing controls contribute to, rather than erase, this perception.

There are models of effective control systems. The worldwide quality movement clearly has proven that control systems with minimal layers and clear roles and responsibilities are far superior to the federal government's system of over-control.(2) In contrast, a control system with multiple layers of supervision, inspection, and other forms of "checking," existing simultaneously in both line management and staff support offices, results in less control and very high costs. Over- control results in less control because those responsible for oversight assume that others also responsible for oversight are performing their functions thoroughly. However, in reality, multiple versions of guidance from multiple groups diffuses accountability and confuses line staffs about what is expected and important.

Fragmented Controls Contribute to an Ambiguity of Responsibility. With so many roles to be played and so many individuals and organizations involved, overlap and confusion of responsibilities has grown. Many strongly differing points of view exist about what roles these groups do play and should play. Clear roles and focus are needed. One premise is that, because line management's own oversight has been allowed to atrophy, staff groups have had to assume these functions. For example, a few Inspectors General now have inspection offices in addition to their standard audit and investigation offices. These inspection offices vary widely in purpose, staffing, and operations, but they do have some common features. They concentrate on shorter-term studies, producing more timely results. These reviews range from quick responses to congressional inquiries, to on-site management reviews, to in-depth evaluations of program practices. These latter evaluations concentrate on program effectiveness and efficiency (how well and at what cost programs are achieving their objectives) and program integrity (how well programs are protected from wrongdoing). GAO on occasion has conducted the same kinds of evaluations, sometimes for the same reasons.

Lamentably, this activity by some parts of the audit community reflects the type of program evaluations some managers believe line management should be conducting internally. It is better in the long term to insist that line managers properly execute their control and oversight functions. To foster this, external overseers should reorient their focus to evaluating the effectiveness of the control and oversight systems put in place by management.

Inappropriate Focus.

Most staff controls are focused on "before the fact" control mechanisms on inputs, such as controls over outlay budgets and staffing levels, and compliance with rules. Most high-performing organizations have shifted their focus to "after the fact" controls on outcomes, such as results-oriented budgets and performance measures. Studies of organizations that have made the shift show decreases in the costs of delivering services, largely because of the ability to reduce the cost of the control systems. While there are some organizations where "before the fact" controls may be appropriate, there are many where a change in approach can improve services and cut costs.(3)

Costs of This Approach to Control are Too High.

The existing management controls system swallows untold millions of staff hours and requires an estimated 660,000 "control" employees at a cost of approximately $35 billion dollars. This estimate includes all line and staff managers and employees performing appraisals, audits, reviews, investigations, evaluations, inspections, and monitoring.(4) It does not include the costs imposed on those that are being appraised, audited, reviewed, investigated, evaluated, inspected, or monitored.

There is no upper limit imposed on how much "management control" is imposed on a given organizational unit or program. Formal audits, reviews, investigations, and inspections in a single governmental organization can easily total hundreds per year and can be initiated at any hierarchical level in the system, from Congress, the central management agencies, or a department headquarters on down. And there are no accounting systems to adequately document these costs. Notably, no one is assigned responsibility for controlling these costs or ensuring these resources are being spent wisely. There is no control over the controls.

There is an alternative that seems to have proven successful in the private sector. The quality management approach--which focuses on preventive actions and results-oriented measures (e.g., customer satisfaction) --can provide a systematic framework for gaining control over the controls. It can help sort out the proper functions of doing vs. monitoring work, of line vs. staff, of internal vs. external control, and of the use of inspection vs. prevention methods.


Redesign the existing set of control mechanisms for the executive branch, using a systems design approach. (2)

The President's Management Council (PMC) should create a streamlined and cost-effective management control system for the federal government. It should begin by documenting the existing system of requirements for control (laws, regulations, rules), classifying the types of methods used (audit, reviews, evaluation), and identifying who does them. It should then assess alternative approaches to use as benchmarks and design a new vision of management control. The PMC should then use this information to design a new, more cost-effective system on behalf of the American taxpayers.

The new management control system should reflect four characteristics:

1. It should be based on systems thinking. Subsystems, such as audits, inspections, investigations, reviews, appraisals, and evaluations, are part of the overall management system. These subsystems must be better defined and their interrelationships better understood in order to eliminate costly redundancy and confusion over accountability.

2. Line managers must be accountable for, and have authority over, management processes and systems. While this is an axiom in high- performing organizations, it is uncharacteristic of today's federal government. The new management control system would:

--- Have line management primarily responsible for management control and be given the necessary resources to accomplish this (e.g., program evaluation staff).

--- Have staff organizations performing a secondary support role for management control. Staff oversight would be minimized, though sufficient to ensure that line managers perform their control functions (i.e., by reviewing their control and improvement systems). Staff should not normally perform control functions for line managers (unless fraudulent activity is suspected).

--- Shift the emphasis from compliance audits, performed by external staff organizations, to ongoing reviews and monitoring performed by line management. Oversight activities must be integral to line management's overall management control system.

3. It should be based on trust, not mistrust. The current approach to management control and staff oversight emphasizes the negative, causing employees to resist management controls rather than use them as guides for improving. Management in high-performing organizations assumes that most people, under most circumstances, are honest, well-intentioned, and eager to contribute to the organization's goals. The small minority who are not well-intentioned should be dealt with accordingly. The government's current approach implies that all employees are suspect; this does not create an atmosphere for constructive improvement.

4. It should be based on measurement and feedback. According to private sector quality management experts, "The right measures are critical to the effective management of work process. If you cannot measure your results, then you cannot control your process and improve your performance."(5) A key principle of measurement and feedback is that people respond to what is measured. Redirecting government from an emphasis on input measurement toward measurement of outcomes, results, and customer satisfAction will be a major step toward this objective.

A second private sector measurement principle is based on the fact that there is variation in all processes. As a result, managers must properly analyze and interpret the meaning of their production data.(6)

The final measurement principle recognizes that waste is present in all processes to varying degrees, but management and employees must continually reduce waste.(7) Some Baldrige Quality Award winners have succeeded in greatly reducing waste. For example, thanks to its "Six Sigma Program," Motorola now has some manufacturing processes that produce less than 3.2 defects per million potential defects.

Cross References to Other NPR Accompanying Reports

Creating Quality Leadership and Management, QUAL02: Improve Government Performance Through Strategic and Quality Management.


1. Senator William Roth, Congressional Record, Vol. 138, No. 51 (April 7, 1992).

2. Deming, W. Edwards, The New Economics (Cambridge, MA: Massachusetts Institute of Technology, 1993).

3. Thompson, Fred, "Matching Responsibility with Tactics: Administrative Controls and Modern Government," Public Administration Review, Vol. 53, No. 4 (July/August 1993), p. 303.

4. See NPR Accompanying Report Transforming Organizational Structures for additional information.

5. AT&T Quality Steering Committee, "Process Quality Management & Improvement Guidelines," AT&T Bell Laboratories, 1987, p. 33.

6. Shewhart, Walter A., Economic Control of Quality Manufactured Product (New York: D. Van Nostrand Company, Inc., 1931).

7. Scherkenbach, William W., The Deming Route to Quality and Productivity (Rockville, MD: Mercury Press/Fairchild Publications, 1986), pp. 40-46.

Return To Report Index

Who We Are |||Latest Additions |||Initiatives |||Customer Service |||News Room |||Accomplishments |||Awards |||"How To" Tools |||Library |||Web Links

Reinvention Comments
Technical Comments