The Heritage Foundation

Backgrounder #1050

August 24, 1995

August 24, 1995 | Backgrounder on

Making Government Work: How Congress Can Really Reinvent Government


The Clinton Administration is making federal management worse. Congress can, and should, do a better job, making employee performance and management accountability the key features of the federal civil service.

When the Administration made "reinventing government" the slogan for its efforts to overhaul the federal civil service, its principal goal was "to make the entire federal government both less expensive and more efficient, and to change the culture of our national bureaucracy away from complacency and entitlement toward initiative and empowerment."1 But the details of the Clinton personnel reform policies and proposals show that this reinvention will weaken management, lead to greater inefficiencies, dilute individual performance initiative, reinforce the culture of complacency, and make government less accountable to federal taxpayers.2

Specifically, a close examination shows that the proposed reforms would:

  • Weaken policies on hiring, promoting, and rewarding federal employees based on individual merit and performance.
  • Eliminate presidential standards for job performance, reducing the President's authority to direct agencies and weakening effective control over the executive branch.
  • Enhance federal agency and union control over setting salary levels, leading inevitably to abuse.
  • Weaken long-standing "veterans' preference" standards for men and women who have served America in the armed services.
  • Create a new role for federal unions that would undermine management authority for defining and assigning work and lead to a more irresponsible federal bureaucracy.
Rather than adopt these regressive proposals, Congress should take the initiative and really reinvent government using principles of sound management. That means Congress must begin by defining the core responsibilities of the federal government and then act to strengthen management systems at the federal level for the remaining functions. As management authority Peter Drucker observed in a February 1995 Atlantic Monthly article, the most important part of "really" reinventing government is to define its proper functions. Specifically, Congress should:
  • Devolve power to the states. Congress should decide which of the federal government's existing functions should be retained and devolve the rest to state governments or the private sector. If new Members are serious about revitalizing the Tenth Amendment to the Constitution, which explicitly reserves all non-delegated powers "to the States respectively, or to the people," they can do nothing less.
  • Define agency missions. For each function retained for direct supervision by the federal government, Congress must specify a mission that is simple enough to be performed in a rational manner with the resources available.
  • Privatize government functions. For these remaining functions, the existing Office of Management and Budget (OMB) A-76 contracting system should be revised so that agencies can provide plans to compete against private bids for relevant agency work under a new competition procedure.
  • Protect the merit system. The Office of Personnel Management (OPM) should prepare an A-76 competition plan for its own operations, privatizing but also ensuring its ability to provide leadership and protect the merit system, and then submit its proposals to Congress for review. In the meantime, Congress should repeal the Ramspeck Act, which allows congressional staff to circumvent the competitive hiring process and burrow into the civil service.
  • Establish responsibility. OPM should create a model "core-spoke-rim" organization to provide guidance in agency implementation, including effective contract supervision and management leadership, and should report to the President regarding agency progress toward this goal. Moreover, Congress should refrain from cutting the number of presidential appointees in the federal government, thus undermining the President's authority to appoint his own people to carry out his agenda. Political appointees, not career civil servants, should be held accountable for Administration policy.
  • Clarify rules. OPM should be ordered immediately to produce management guidance for an attrition-first government personnel reduction policy, including a possible personnel freeze and procedures to protect the pension system from any abuse of early retirements and buyouts.
  • Re-establish merit selection. OPM should be ordered to revise the Uniform Guidelines for Employee Selection Procedure to comply with the law on equal opportunity and merit selection. It also should be required to submit these revised guidelines to the court and to revise all examinations to ensure compliance.
  • Make hiring rational. OPM should maintain central and open administration of common-function occupations for efficient hiring, in addition to overseeing classification and performance management while continuing the general Carter-Reagan policy of decentralized management and centralized oversight.
  • Re-establish pay for performance. Congress should reinstitute a management-directed pay-for-performance system -- this time for all federal employees and focused on contract management -- that preserves the Carter-Reagan system of responsible political and career executive direction and control, as well as appropriate employee consultation.
  • Rationalize federal benefits. OPM should be ordered to prepare a series of options for reform of the government's compensation structure, including making pension and other benefits more portable and competitive.
  • End duplication. OPM should be ordered to prepare a plan to eliminate the duplicate grievance system (insofar as possible), replace it with a single merit system, abolish appropriate supporting institutions, and reassign any affected personnel. It should prepare a plan to simplify the appeals process by consolidating the cases, functions, and staff of the Merit Systems Protection Board, the Federal Labor Relations Authority, and the federal government oversight responsibilities of the Equal Employment Opportunity Commission.
  • Stop Clinton's devaluation of management responsibility. Congress immediately should overrule President Clinton's Executive Order 12871 devolving essential management responsibilities to labor-management committees. This might be accomplished most simply through an appropriations rider.
  • Consolidate the management of federal resources. OPM should be ordered to prepare a plan to consolidate the Office of Management and Budget (OMB), OPM, and the General Services Administration into a new Office of Management, and to submit this plan to Congress.
Making Government Accountable to Federal Unions

The Clinton Administration began its "reinventing government" campaign by announcing that it had cut thousands of pages of red tape by dismantling the Federal Personnel Manual (FPM). The FPM was a repository of laws, rules, and management guidance for the federal workforce, put in one place for easy reference at low cost. Now even Congress cannot obtain a complete copy of all personnel regulations and guidance. Personnel officers confide that they have hidden their old copies for personal, presumably secret reference. But what happens when this generation of management experts is gone? Despite all the media hype, this most publicized Clinton "reform" is merely a revealing example of counterproductive news-release management.

"Reform" is no guarantee of improvement. Reorganization can, and often does, make an organization weaker and less efficient. A Cabinet agency can be eliminated simply by shifting its functions to another department, but the result also can be even more ruinous, confusing, and costly regulation. Similarly, downsizing can save or cost the taxpayer money, depending on how much is contracted out and how many personnel are separated -- and on what terms. And while devolving functions to local governments usually is more efficient, decentralization cannot be depended upon to have this effect if it merely buries decision-making deeper within a single large bureaucracy without any other rationale.

For President Clinton, "reinventing government" presented a serious political problem from the outset. The Administration wished to make government more efficient and reduce personnel, but it had relied on the federal employee unions to get elected and could not afford to alienate them.3 It was clear that these unions could not support, or even ignore, a proposal to cut 252,000 personnel slots unless they received something very valuable in return.

Contrary to the Administration's own staff recommendations, which were overruled personally by Vice President Al Gore, the White House decided to give the unions equal power with management in "labor-management councils" that would make the major management decisions in agencies of the federal government. In addition, it was proposed originally that the unions be given an involuntary dues checkoff from federal employees -- without even a requirement for representation elections. While the White House was forced to retreat from the second proposal, the first was codified in Executive Order 12871, issued in 1993, making the unions "full partners" with management in the assignment and classification of work and creating labor-management committees to enforce this throughout the government. A presidential "partnership council" of union and Administration officials was created to make further recommendations, including the proposal for involuntary dues collection and union representation by card submission rather than by secret ballot.

Following the 1994 election, the White House resurrected the council's recommendations as part of a "New Democrat" initiative to reform the bureaucracy. A leaked draft proposed removing "dead wood" personnel and giving agencies more "flexibility" over job classification, pay, hiring, and performance management decisions to increase efficiency. No doubt in recognition of the new Republican majorities in Congress, the involuntary union dues and card-submission plans were shelved and the more appealing proposals highlighted.

Clinton's Plan to Reinvent Government

In January 1995, the Clinton Administration released "draft specifications" for proposed legislation to implement the National Performance Review (NPR). The draft generally followed the National Partnership Council's recommendations to reinvent the "Federal Government's human resource management (HRM) systems and processes." Final recommendations issued in May were diluted further in an effort to prevent dismissal of their core elements by a Congress now controlled by the other political party.

The Administration's recommendations certainly would "reinvent" government as taxpayers have known it. But would they make government be more or less efficient and accountable than it is now? The following discussion analyzes the executive order (the only part actually implemented) and both the draft and final recommendations, the latter because they are now before Congress and the former because they presumably still represent what President Clinton would submit to a Democratic Congress in the event one is elected in 1996 and he is still in the White House.

Undermining the Hiring System

The law on hiring in government is precise: "selection and advancement should be determined solely on the basis of relative ability, knowledge and skill, after fair and open competition which assures that all receive equal opportunity."

While this rigorous legal standard is often skirted in practice, the Clinton recommendations would legitimize present questionable "examinations" and encourage new ones that improperly use other considerations. Agencies would have the explicit power to appoint individuals non-competitively to any positions designated by the Director of the Office of Personnel Management as "shortage" positions or as requiring "exceptional qualifications."

Government unions long have sought to bargain over so-called crediting plans and to replace legally established knowledge, skills, and ability criteria (KSAs) with seniority or "qualifications" or on-paper educational attainment criteria. Ethnic and gender affirmative action criteria have been used for years to skirt merit selection requirements. Under the Clinton proposals, OPM would set minimum standards -- which the Clinton draft proposals specifically provided and the final ones imply -- that agencies could "augment," allowing these other criteria possibly to dominate the KSAs.

The draft proposals would have allowed agencies "to determine appropriate evaluation methods, announcements, and other processes," and even "when they need to announce jobs," thereby eliminating procedures that allow for open competition. This would effect a long-time goal of the federal union leaders: to give first consideration ("sequential consideration") to members of their own bargaining units rather than allow open competition among all American citizens who apply, as called for under present law.

Most seriously, the final Clinton proposals would provide an alternative ranking process under which candidates for positions could be placed in "quality categories" based on levels of qualifications rather than individual scores. This dilutes relative ability rankings and undermines selection based on individual merit. Veterans also would be placed within these quality groups, with the individual protections to which they are now entitled limited accordingly.

Veterans' preference in hiring and promotion would be diluted further under the recommendation because individuals in other disabled categories, including alcoholism, drug dependence, or mental depression, would be given the same preference as those who have borne the brunt of battle in military service. For lawyers, it would be eliminated entirely. The greatest complaint among veterans organizations is that federal agencies do not follow the OPM regulations on preference closely enough. This situation would be aggravated, not helped, by devolving the examination process further to the agencies and placing individuals in group categories.

A second set of proposals under the hiring section would reclassify temporary appointments as "nonpermanent." Such employees would be placed under union grievance procedures (after one year), would be allowed an advantage in competing with outside applicants (after two years), and would receive fully paid health insurance and retirement coverage (after a year) and within-grade increases (as earned). The practical effect would be to remove many of the flexibilities and lower cost aspects of temporary employment and make it less attractive to hire temporaries.

The stated purpose of hiring reform is to streamline the government's "highly centralized" system. The recommendations are portrayed as decentralizing authority to agencies. But hiring for 85 percent (over 700) of the government's examinations already is decentralized to the agencies, except for so-called common-function appointments. The final recommendations would allow these also to be decentralized even though they are less costly, and can be fairer and more effective in determining merit, when administered by a central authority.

The Clinton hiring initiatives thus would promote the appointment of individuals to positions without the competitive KSA examinations required by law, would weaken individual merit as a principle in civil service appointments by grouping candidates into "quality categories," would erode the existing special appointment status for veterans, and would limit the flexibility of temporary appointments. Some other technical recommendations could be enacted -- including one extending the probationary period up to three years -- but the proposed reforms hardly would reinvent merit hiring in any positive sense.

Decentralizing Classification Authority

The Clinton reforms would give OPM, federal agencies, and unions substantial control over pay, reducing the authority of Congress over this critical personnel matter. OPM would be directed to establish its own criteria for pay classification, thereby abolishing any statutory definition or control. OPM would set criteria to "broadband" existing grades, but agencies could implement their own systems without prior approval from any central source -- except, apparently, a labor union through the labor-management councils established by the executive order. On the positive side, the recommendations would make currently automatic "periodic step" (within-grade) and "additional step" (quality step) increases contingent upon performance, although agencies would base their individual pay decisions on performance appraisals developed by the labor-management councils.

OPM's power to revoke agency classification authority when abused by management would be repealed, as would the requirement that it review the classification of positions, except when such review is requested by the employee. In other words, unless OPM itself desired to undertake a review, it probably would be done only to increase the pay grade of an employee, since no one normally complains about receiving too high a salary. OPM could still review systems for compliance -- but only to direct corrections, not to force them. With OPM relieved of the obligation to assure quality, there would be less supervision of compensation and a vastly increased potential for abuse.

The practical effect of the Clinton classification proposals would be to transfer authority over pay increases (other than across-the-board general pay) from Congress and the President to agencies and their unions. Each agency undoubtedly would create expanded internal reviews and additional control systems, thereby also creating a need for more staff. The President, OPM, Congress, and the taxpayer would be asked to trust the agencies with precisely those decisions it is most in their interest to abuse.

Eroding Performance Management

All presidential or other government-wide standards of job performance would be eliminated under the Administration's plan and replaced with performance standards bargained between agency management and unions. Collective bargaining would be imposed not only for the design of work plans for all employees, but also for setting what work is expected of each individual, determining whether those standards are met, and establishing how the work is to be assigned. Agencies would be required to create award programs to provide incentives for individual or group achievements, with the nature of the awards process also determined by collective bargaining.

The proposed reform most heralded by the Clinton Administration would allow agencies temporarily to reduce the pay of poor performers for a period not to exceed 120 days. The hidden agenda lies in an innocuous-sounding proposal to eliminate "dual track" actions against poor performers, ending performance removals (under Chapter 43) and allowing only disciplinary-action removals (under Chapter 75). This one change would vitiate the performance management reforms of the Civil Service Reform Act of 1978, the whole idea of which was to introduce a fair but management-determined procedure (subject to appeal) to allow removal or denial of pay increases for poor performance without having to resort to more draconian, demanding, and cumbersome procedures (including labor grievances and appeals).

This reform would undermine the ability of the President or his representatives to manage the executive branch. If they cannot set broad rules, they cannot set general management policy. The proposed change would not return to the status quo ante. It would impose collective bargaining not only over the standards of work to be performed by employees -- which some might accept as reasonable -- but also over the performance of the work itself and all disciplinary procedures under it. It would realize in law the promise President Clinton made to the labor leaders in his executive order: that unions would be "full partners" with career managers in managing the federal government. For all practical purposes, the President, his top appointees, and Congress would be relegated to the role of outside observers.

Weakening Management Rights

The number one goal of federal union leaders has been to weaken the strong "management rights" section of President Jimmy Carter's Civil Service Reform Act of 1978. The Clinton proposal would eliminate both the "permissive" right to decide appropriate staffing levels and the corollary right to decide how work is to be performed. The executive order simply directs management not to exercise this right; the proposed change in law would eliminate it as an option for any future President.

The management rights provision of current law would be replaced with a "good government standard" under which agencies are "obligated to bargain collectively" over how all work is to be performed in the federal government (including much of Congress). The executive order would establish labor-management "partnerships" throughout government, as well as a government-wide National Partnership Council, and would create an agency-level partnership council of management and unions "to develop agency policies and regulations affecting conditions of employment that are binding on agency components and bargaining units subordinate to the council." Under the draft, appeals to any "statutory third party" would be disallowed; decisions of the joint labor-management councils would be final.

Clearly, under the Clinton proposals, labor-management councils would make the major management decisions. Management, even for such essential functions as how many employees are to perform work, how they are to be compensated, and how the work is to be performed, would be by committee. It is difficult to see how this would improve efficiency.

Limiting Presidential, Congressional, and OPM Oversight

The role of the President and his managerial right arm, OPM, under Clinton's proposed reforms is ambiguous. While one of the principal stated proposals is decentralization of management authority from OPM, the agency would retain some type of oversight authority. Indeed, under the alternative personnel systems proposal, OPM alone would have the power to approve even systems radically different from the one contemplated by the proposed new law. It would be obliged only to "notify" Congress of departures from the present proposals, although any change would have to be requested first by a labor-management partnership agreement.

Under the Clinton proposals, OPM could order corrective classification action but no longer could revoke the authority. In other words, an agency could continue classifying while not in compliance. OPM's only enforcement power would be informing the President of violations "involving agency heads and directing corrective action." Thus, while the effective power for personnel decisions would be exercised by labor-management committees, the responsibility would still rest with the political appointee heading the agency, enforced by the President.

The President's OPM Director could try to persuade or dissuade agency heads, but none of them would have the power to overrule the labor-management committees. Nor would OPM, the President, or (under the draft) any other "statutory third party." The only enforcement would be for the Chief Executive to remove his agency head from office.

Unable to order agency compliance, or even to set prior presidential standards for performance, the OPM Director would be ignored -- except by an annoyed President, notified when the abuse reached the level of a political crisis involving the resignation of a top official, or by an aroused Congress looking for a scapegoat. Neither Congress nor the President would be able to hold responsible the labor-management committees making the real decisions.

How to Achieve Real Reform of Government Management

A reform supposedly meant to create "high performance government" thus would permit unions and management councils to make the major decisions without having to shoulder any responsibility -- and at the expense of those with legal responsibility for performance and oversight. Instead of reinventing government, Clinton's proposal would make things worse by transferring leadership to committees and unions and assuring weak executive leadership. Management expert Peter Drucker argued as early as 1985 that building "entrepreneurial management" into public institutions may be "the foremost political task of this generation." The Administration's proposal ignores this challenge to improve entrepreneurial performance. Instead, it builds more bureaucracy.

Given the strong public support for more efficient and more accountable government, it would be a breach of Congress's oversight responsibility to accept President Clinton's radical plan simply to appease union leaders. The Clinton reform would make the task of managing the bureaucracy even more difficult. Members of Congress therefore must provide an alternative that really reinvents government. To do so, however, they must master the details of bureaucratic administration, especially when "reformers" use terms in ways that confound common understanding. For example, "decentralization" makes perfect sense in the private sector because the profit-and-loss bottom line clearly signals whether devolution has gone too far, endangering profitability or even survival. Devolving functions to state and local governments also makes sense because local voters can sort out the details more easily and fire elected lower-level officials if things go wrong.

But decentralizing personnel management decisions usually just shifts them further away from executive and congressional oversight. Generally, it creates less responsibility, not more. Labor-management councils, quality circles, and group management ideas can be used in government, but only in limited ways and far less than is possible in the private sector. These methods can increase quality, but they have the offsetting cost of delaying decisions. Only the knowledge that profits are threatened forces a decision in the private sector, and that decision is made ultimately by top management, not a committee. In government, with no profit mechanism to limit delays, a culture of inertia becomes the rule.

Back to Basics: The Principles of Real GOvernment Reform

Government is very different from the private sector. In his classic work Bureaucracy, the great Austrian economist Ludwig von Mises showed that government management is more difficult precisely because government lacks the price system and bottom line that force action in a market.4 Prices can simplify complex problems by reducing them to one overriding question: Is the unit making a profit? But government has no such simplifying device. It is therefore essential to understand the different management principles required for government administration.

Principle #1: Keep the mission simple.

In 1971, responding to desperate pleas for assistance following massive loss of life and property in a series of hurricanes in the 1940s, the Army Corps of Engineers had turned 103 miles of meandering, mosquito-infested Florida swamp, the Kissimmee River, into a 56-mile canal system that protected local citizens from floods by utilizing the most modern electronically operated locks available to engineering science.

Eleven years later, in 1992, Senator Bob Graham (D-FL) threatened President George Bush with the loss of Florida's support if he did not agree to re-route the Kissimmee canal back to the original "river" at a cost of almost half a billion dollars. And who was to do this? The same Corps of Engineers, which understandably did not like the idea of destroying its masterwork canal and turning it back into a swamp.

1971 was the era of "can-do government," and the Army Corps of Engineers was the government's best. Only the Marine Corps, Federal Bureau of Investigation, Social Security Administration, and Forest Service were in the same league. All had mastered what James Q. Wilson, professor of management at Harvard University, would identify as the essential ingredient for administrative success in government: a simple, clear mission with an imaginative leadership dedicated to that mission. Government works if it keeps its mission simple and continuously drives that simple message home to a cadre of dedicated employees.

Each of these agencies had an unambiguous sense of mission because its founder made certain that his legacy was ingrained in his troops and communicated to all new recruits. It was, as Wilson stated in his book Bureaucracy, "as if they felt the ghosts of Sylvanus Thayer [Corps of Engineers], Arthur Altmeyer [Social Security Administration] and Gifford Pinchot [Forest Service] looking over their shoulders."5 Like the other two --J. Edgar Hoover (FBI) and Commandant John Russell (Marine Corps) -- they made government work.

In the intervening two decades, government missions have become enormously more complex. What made the Corps of Engineers efficient was the ability of its single-minded engineering mission to withstand changes in the political wind for over half a century. What made it great was its pride in creating engineering masterworks. When the political fashion changed from dams, hydroelectric power, canals, and flood protection to environmentalism, it was clear that the mission would have to change. But good government management requires simplicity of mission, and the Corps is cracking under the ambivalence created by trying to absorb environmentalism into its engineering ethos.

The same is happening to the other elite agencies. The Social Security Administration (SSA) was unmatched when it processed claims for elderly Americans; when Congress gave it supervision over disability, its "pay benefits on time and accurately" ethos broke down. Deciding how old one was and whether one had contributed to Social Security were simple matters, but evaluating medical evidence was often subjective and very difficult. Giving checks to some and not to others seemed unfair in the SSA culture.

Government foresters were fine at managing their domain until economists, engineers, and conservationists were forced into the Forest Service by Congress, bringing with them different definitions of "forest yield" based on their own conceptions of "good" forest management. The result: divisions were created that confused the Forest Service's mission and restricted its performance.

Even the FBI's ethos of clean, professional, and straightforward investigation was transformed as the Bureau was pushed more into drug, mob, and gang investigations, all of which require difficult undercover work and unorthodox methods. The result: the bloodshed at Waco and Ruby Ridge.

The paradox is that liberals in Congress and elsewhere who wish government to do all good things are the ones who destroy its ability to do so. Government can work efficiently only when relatively few things are assigned unambiguously to a few institutions with sufficient esprit de corps to do them well. Because the Founders understood this, they created a system in which responsibility was divided between levels and branches of government, and between private and public sectors, so that the work of society could be administered more effectively.

Principle #2: Keep national functions national.

American government has strayed far from its federalist roots. Beginning with President Woodrow Wilson's Fair Deal, and during World War I, an expert federal bureaucracy was created that has continued to expand its powers. A temporary weakness in state government and business finances during the Depression era of the 1930s led to President Franklin D. Roosevelt's New Deal and its dramatic expansion of the national government's position as the preeminent sector of American society. With President Lyndon B. Johnson's Great Society in the 1960s, the national government was spending almost one-quarter of the nation's total wealth and exercising regulatory control over almost every aspect of the economy and society.

Even before the 1994 election, a serious public reaction was developing. Veteran Washington Post columnist David Broder noted that "federalism issues are back on the national agenda in a serious way." Governors were bargaining with the President and congressional leaders in a way not seen in years, if ever before. The original federalist idea was to specify, primarily in Article I, Section 8 of the Constitution, the powers of the national government and to leave the rest to the states or to the people. That idea, specified in the Tenth Amendment, largely disappeared from court decisions after the 1930s but now seems to be undergoing a serious revival that extends into actions of the post-1994 Congress and even into recent court decisions.

The discussion on welfare policy reform, for example, has changed dramatically since the 1994 election: Now both political parties are outbidding each other to show their desire to decentralize power. There are serious proposals to transfer many of the major welfare programs, such as Aid to Families with Dependent Children (AFDC), public housing, and even Medicaid, to the states. Members of Congress, and even the President, are re-evaluating which functions should be performed by the national, state, or local governments and which should be performed privately, based on constitutional grants of power and a rational division of labor.6

Functions properly performed by the states should be transferred back to them, together with a corresponding federal income source. Devolution to the states, and through them to local communities and the private sector, promotes efficiency by simplifying the work. Both the traditional constitutional grant and James Q. Wilson's modern administrative logic suggest that national functions be limited so that they can be performed well. In a complex, interdependent world, it is essential to de-construct large bureaucratic structures, in both the public and the private sectors, to keep them from being overwhelmed by red tape.

Principle #3: Get serious about privatization.

Once the new, streamlined national government's functions are set, a budget can be devised and decisions can be made regarding what should be done by government employees and what should be contracted out to the private sector. Governors and mayors across the United States recognize the reduced costs, greater efficiency, and improved management that flow from contracting work out to the private sector and utilizing other forms of privatization. Heritage Foundation analysts have identified hundreds of billions of dollars in federal assets that could be put to more efficient use in the private sector while increasing federal revenues.7

Still, unlike chief executives at lower levels of government or national leaders in other countries, no recent U.S. President has given contracting out and other forms of privatization the top-level attention they must have in order to prevail against the predictable resistance of bureaucratic interests. The current U.S. program exists as a neglected backwater within the Office of Management and Budget and receives almost no support.

One reason so little is contracted out in Washington is that the OMB A-76 procedure for comparing government and private costs for the same work is skewed toward having the work done by government. Federal pensions, for example, are badly undervalued and understate the cost of government management. The procedure for comparing costs needs to be reformed if it is to be of any real value in deciding how work should be allocated.

Another reason is that not enough attention has been paid to winning, or at least neutralizing, federal employee support. Giving part of the savings from contracting out to managers and employees who recommend privatized services, to employees who agree to shift with a function to the private sector, and to those who remain to oversee the operation, for example, can reduce employee opposition. The more or less moribund FED CO-OP program -- designed to give shares in private firms to federal workers who assist in making the transfer -- should receive greater attention. Indeed, the whole work of contracting out and contract management must be reorganized in a new and more rational manner.

Principle #4: Get serious about performance management.

The new workforce must be organized properly for a contractor-based system. Actually, contract employees are already predominant: perhaps eight million, compared to the federal government's two million civilian employees. Millions of state government employees also implement federal rules. No one knows the total number or is in a position to manage it properly. Government is still organized as if it were the 1930s.

A new "core-spoke-rim" organizational structure must be created. Core federal employees would be expert, highly compensated executives charged with setting plans and managing the contractors who perform most of the work on the rim of government. In between, spoke (temporary) employees with few benefits and little or no job protection would be used in cycles demanding more work than the government's core employees were able to perform.

This is where the Clinton reforms go most astray. Rather than increase managerial flexibility, they would expand the bureaucracy by layering labor-management committees at multiple levels of administration and create more permanent employees by granting protected status to formerly temporary employment. They unnecessarily divide central responsibilities, further encouraging duplication of work within agencies. They also ignore the negative effect on pension flexibility. Even a larger permanent workforce would be more mobile if pensions were made fully portable, like 401(k) plans in the private sector; no employee would have to be tied for 30 years to a job that has become obsolete and that he or she has come to hate.

One of the principal advantages of contract management is that it rewards performance. Rather than leave pay classification and pay-for-performance to the mercies of labor-management committees, OPM should be instructed to transmit to Congress a compensation system that rewards both savings and mission accomplishment that is on time and as specified. Classification should be broad-banded, but only under congressional rules and tight OPM supervision to reduce the normal temptation of agencies (shown in demonstration studies already conducted) to push compensation up unnecessarily. While employee and even union input are essential, the final decisions ought to be made on mission-accomplishment grounds by top agency management under the supervision of the President.

Principle #5: Get serious about merit.

Because a core-and-rim staffing system requires a higher quality permanent employee, it is important to restore selection based on knowledge, skills, and abilities. OPM should seek immediately to end the sweetheart consent decree, entered in the last days of the Carter Administration, that abolished its Professional and Administrative Career Examination (PACE) for competitive selection of superior college graduates and replaced it with a credentials-based system requiring that blacks and Hispanics be hired in proportion to the number standing for the examinations. The decree was to last only five years but already has placed the federal courts in control of hiring under non-merit requirements for 15 years.

There is a reason for centralized hiring. General ability tests like PACE are better and cheaper than any separate tests for particular occupations. Unfortunately, however, the federal courts have ruled that such tests raise the problem of disparate impact against racial, ethnic, or gender groups. While an argument might be made for some temporary form of affirmative action to assist victims of previous discrimination, 15 years without a merit entry examination certainly deserves some notice and redress. The courts have agreed to review the decree if the Uniform Guidelines on Selection Procedures are reformed. This challenge must be accepted so that federal hiring once again comports with the law requiring selection based on knowledge, skills, and ability (KSA) criteria. A central agency like OPM, with the knowledge base to take on the job of updating PACE, should be assigned this responsibility.

Centralized examining for common positions is still required by law. It also is cheaper. OPM has been shown to select employees at between $10 and $15 less per applicant than other agencies. Core-and-rim organization has other advantages. In general, it takes less hiring and fewer personnel resources for a small core of skilled professionals to manage a larger number of temporaries and contractors with a single mission focus.

Just as the law requires that hiring be based on skills, retention and reward are supposed to be based on good performance. The Clinton reward and disciplinary systems are to be operated with substantially greater union involvement. Indeed, the apparent purpose is to shift totally to labor-management control and union grievance review. While intelligent mechanisms for employee involvement can be devised, there is no prospect that a public-sector, union-dominated system like the one envisioned by the Clinton reforms will lead to higher standards of performance and more action against poor performers. Unions do not thrive by being tough on employee performance or discipline.

True labor reform requires an entirely different approach: eliminating the expensive and duplicative grievance system and re-establishing a true merit system. This was the justification for creating a civil service in the first place. It is what President Carter desired before a union-dominated Congress forced him to compromise and create the present system. That abuse of power can be corrected by substituting the Merit Systems Protection Board (MSPB) appeals system as the exclusive remedy. The responsibilities and staff of the Federal Labor Relations Authority and the federal government oversight responsibilities of the Equal Employment Opportunity Commission should be merged into the MSPB to provide a "one-stop" appeals process. OPM should devise a plan to integrate these functions under general merit system principles and then submit this proposal to Congress and the President.

Principle #6: Get serious about management.

President Carter's 1978 reforms gave political appointees the tools they needed to manage the bureaucracy. President Ronald Reagan implemented and advanced these same tools. For a few years, it worked. More was accomplished with less as measures of productivity increased and personnel were cut.8 But the thrust of the Clinton reforms is to remove authority from the political and career executives responsible for better management under the Carter-Reagan reforms and transfer it to labor-management councils. The deliberate weakening of political control is a long-sought goal not only of union leaders, but also of career manager-dominated groups such as the American Society of Public Administration. A recent Brookings Institution study, typical of the Washington bureaucratic mindset, suggests cutting the already minuscule number of political appointee positions by one-third. Incredibly, Congress has incorporated this proposal in its budget resolution, thus advancing the political agenda of Washington's bureaucratic establishment.9

Without political leadership, when the President gives an order, there is no reason to assume anything will happen down the line in the federal bureaucracy. That is why responsibility in current law rests with the political agency head. In a perverse way, the Clinton recommendations acknowledge this by proposing to transfer power to labor-management councils. The recourse against abuse, however, is the political agency head. Democratic government, at some level, must place responsibility in the hands of political appointees representing the elected executive.

The Carter Administration's management reforms recognized this essential reality and moved responsibility down the management chain -- through successively lower levels of political executives, career executives, and managers -- to where the work was performed. All was bound together with a management-directed performance appraisal and reward system to implement the policy set by the President under the laws of Congress. Employee work groups and organizations can be useful in some situations, and managers should consider employee input and needs. But effective government management requires strong agency leadership in the tradition of Thayer, Altmeyer, Pinchot, Hoover, and Russell, backed by effective oversight. Though these men came from the career ranks, they were effective in political positions with the support and supervision of the President. The more the mission is simplified, and the work contracted out and rationally organized around performance criteria, the more such leadership is possible.

It is expecting too much of subordinate career executives and union leaders to make pay, hiring, firing, and performance-rating decisions independent of political executives representing the President. Only political appointees -- because their rewards come from the President, not the career civil service system -- have any incentive to resist the dominant cultural pressures on management not to make tough decisions.10 Turning control of these executive decisions over to unions makes even less sense. In an environment in which the public is demanding less but more efficient government, their whole purpose remains getting more government positions at higher compensation for their members.

The Clinton proposals would make things worse. The solution is to return to the Carter-Reagan reforms and make them work; and they can work with intelligent and dedicated leadership operating under sound management principles.

Managing the Government: The Function of OPM

Whatever direction management takes in the future, it is clear that the federal government will be smaller and that personnel will be reduced. President Clinton recommended a reduction of 252,000 positions, later increased to 292,000 in order to fully fund the 1994 crime bill. New budget plans continue this trend. The Clinton approach is for agencies to set their own personnel plans. The alternative is to have a presidentially directed and agency-implemented plan to manage the process efficiently, adhering to sound human resources management principles. This is the purpose of the Office of Personnel Management: to help develop and to manage government-wide personnel priorities set by the President.

As Chief Executive, the President has the constitutional responsibility to "take care that the laws be faithfully executed" and is given subordinate officers to assist in that task. One of these officers is the Director of OPM, responsible for "administering and enforcing the civil service rules and regulations of the President." Most of OPM's work could be (and should be) contracted out -- but not its job of helping the President manage the vast federal bureaucracy.

By forbidding the Director or the President from issuing government-wide performance standards, the Clinton proposals undermine the President's constitutional right to manage the executive branch. By transferring management rights from agency political heads, the OPM Director, and the President to labor-management councils, they strike at constitutionally protected executive responsibilities. By granting agency labor-management committees the right to design and administer pay classification, testing and hiring, work allocation, and performance management systems, they subvert the authority of Congress and the President to set basic management policies for the administration of government. They would deny the Director of OPM the tools needed to assist the President in managing the government. Decentralization to agencies should be the operational model, but it is essential to retain some central direction and oversight by the President and his associates, such as the Director of OPM.

It is irresponsible for the executive branch not to devise the outline of a central plan to manage the proposed personnel reductions. OPM -- presumably still possessing the necessary human resources management skills -- should be a central part of that responsibility. It need not micromanage. Without incurring major costs or inhibiting agency flexibility, a plan could be effected humanely by basing personnel reductions on "attrition first," as was done in the early Reagan years. If needed, a total freeze should be set early, with exceptions only for critical skills and essential functions. To keep agencies from subverting this process, it must be monitored centrally by an expert, clearly focused personnel agency led by an official strongly committed to the President's personnel reduction goals.

Some might expect the Office of Management and Budget, within the Executive Office of the President, to manage these personnel reductions. But OMB has neither the special skills nor the clear focus essential to this mission. Because the budget, not personnel management, necessarily dominates its perspective, OMB analyzes the cost of personnel, not needed staffing levels and skills, and not with the overall objective of reducing bureaucracy and increasing efficiency.

The Office of Personnel Management, by contrast, can devise a government-wide plan. It has -- or can regain -- the knowledge of agency operations needed to assess true requirements. Through its special pay rates program, it can determine the need for specialized skills and provide the means to secure them. Thus, under a modified personnel freeze, OPM could evaluate agency requests for exceptions.

The best answer, however, and one that could save substantial funds, would be to merge OMB, OPM, and the General Services Administration (GSA) into a single Office of Management (OOM), as recommended by the Grace Commission and other outside evaluations. This also would allow the contracting-out function to receive appropriate attention.

Reductions in force (RIFs) and furloughs should be utilized only when needed to keep costs within budget or to assure lower long-term costs. If RIFs are used at all -- as they will have to be if certain operations are restructured -- the government should adopt the Reagan Administration proposals and base employee retention during downsizing more on performance than on the current seniority-dominated weighting process. It also should limit so-called "bump-and-retreat" rights, under which employees are "bumped" out of the service by higher-level individuals with greater seniority who are overqualified (and overpaid) for the lower-level positions.

Modifying existing practice not only would cause better workers to be rewarded and the work product to be upgraded, but also would mean that women and minorities were not affected disproportionately affected by RIFs simply because they tend to have the least seniority. Nor do so-called buyouts make sense. Giving an employee $25,000 to retire, in addition to what it costs at the time, inordinately increases the cost to the retirement system. Ultimately, increased costs lead to more separations than necessary. Moreover, a "core" employment emphasis relies on an expert workforce, and buyouts induce the most skilled to retire.

By following the Reagan Administration's guidance, emphasizing attrition while allowing some RIFs and furloughs, it should be possible to minimize the negative effects. Over 90 percent of the Reagan reductions were achieved by attrition -- and many of those affected were moved to other positions through a newly instituted placement program. By contrast, during the Eisenhower Administration -- the only other recent administration to reduce the bureaucracy by any comparable size -- almost 90 percent of the cuts were achieved by firings.

Any rational reduction in personnel requires a plan that is both flexible and comprehensive. Administrative savings, though often abused as a solution, also must be part of any overall plan. An "attrition first" policy is the first step toward an efficient solution. Benefits also must come under budget scrutiny in this era of limited resources. For example, federal retirement represents four percent of the budget going to relatively few people who, while they should not be unfairly disadvantaged, enjoy earlier retirement and more generous benefits than all but a handful of their fellow citizens.

Without reasonable reform plans, unreasonable ones will be imposed. Fortunately, reasonable approaches do exist. A relatively modest pension reform, for example, could limit future cost-of-living increases to the maximum dollar amount of the Social Security COLA increase. In addition to being more equitable across different federal pension plans, this could save $20 billion over five years. Increasing the retirement age past 55 over a period of time would save many billions more. Moreover, these changes might well be the least disruptive to the workforce and certainly could be implemented for new employees. Other approaches may be better, but an agency like OPM should assess them from a government-wide perspective if anything rational is to be accomplished.

Taxpayers are demanding real reform, and personnel policy must be part of the necessary re-evaluation of government. Total personnel costs equal 13 percent of the budget, and overhead adds another 4 percent. This is not the time to decentralize to labor committees and hope that all turns out well, as the Clinton reforms would do. Only an aggressive, centrally managed policy can insure that real reform takes place.


The national government badly needs reinvention. The career managers and public administration associations would return to the pre-Carter system of decentralized authority in the hands of career managers. The unions and the Clinton Administration propose to decentralize authority to labor-management committees. But the best solution would be similar to the Carter-Reagan reforms -- placing management authority and responsibility in the hands of the elected President and his appointed and career team, limited by the protections of an enforced merit system. Unfortunately, because some of the essentials of the Carter reforms were compromised to assure passage, the government has two personnel systems operating together. A true reform would complete the reorientation toward performance management begun under Presidents Carter and Reagan by ending this anomaly.

President Clinton issued Executive Order 12871 to implement part of his reforms of the civil service by decree, primarily by ordering agency heads to ignore management rights in labor negotiations and creating labor-management partnership councils "as full partners with management" in areas not forbidden by law. Even with its powers limited by its refusal to change the law, Congress will be able to see how well these changes work. So will the taxpayer. The presidential veto and Mr. Clinton's desire to satisfy the unions probably make any alternative reform most unlikely under the present Administration. Setting out the requirements now, however, will help prepare the way for future reform.

An expert, core workforce will demand merit system hiring, based on skills and open to all, with reasonable preference for those who have served their country honorably in the military. Staffing will need to be flexible: a small, permanent contract-managing core, temporary employment fluctuating with demand, and most of the work performed by contractors. Job classification should be made more flexible but must be monitored carefully. Performance should be evaluated under mutually defined but managerially decided standards, directed and run by strong executives pursuing a simply defined mission and responsible to presidentially designated agency leaders, all subject to oversight and review.

President Carter tried to establish such a system in the Civil Service Reform Act of 1978, and President Reagan implemented its performance-based system in 1981. This included decentralizing most management and examining functions to the agencies and their responsible chief executives, where they now reside, under the general supervision of the President. The Clinton reforms would rely on labor-management committees and decentralize further, undermining presidential leadership and efficient management at great taxpayer cost.

So far, Clinton's emphasis on reinventing government has engendered favorable media treatment of proposals that would make matters even worse than they now are. The challenge to Congress is to "reinvent" government effectively by devolving functions to local governments and the private sector and by intelligently restructuring and reforming the remaining federal functions so that they are conducted more efficiently and at lower cost.

It cannot be completed overnight, but true reform must begin now. A people demanding smaller and more cost-effective government can accept no less.


  1. For an early assessment of the Clinton effort at government reinvention, see Donald J. Devine, "How To Cut the Federal Bureaucracy," Memo to President-Elect Clinton No. 2, December 14, 1992. See also Adam D. Thierer, "The National Performance Review: Falling Short of Real Government Reform," Heritage Foundation Backgrounder No. 962, October 7, 1993, and Donald J. Devine, "Why President Clinton's Reinvention of Government Is Not Working," Heritage Foundation Backgrounder No. 970, December 28, 1993.
  2. The corpus of the Clinton Administration's civil service proposals can be found in three documents developed by the United States Office of Personnel Management (OPM): "Human Resource Management Reinvention Act of 1995" (proposed legislation), May 24, 1995; "Draft Specifications for HRM Reform Legislation," January 20, 1995; and "Report of the Federal Labor Law Reform Working Group," January 8, 1994.
  3. For example, the Clinton Administration supported the repeal of key provisions of the Hatch Act, which prohibited civil servants from getting actively involved in partisan politics. See Robert E. Moffit, "Gutting the Hatch Act: Congress's Plan to Re-Politicize the Civil Service," Heritage Foundation Issue Bulletin No. 180, July 6, 1993.
  4. Ludwig von Mises, Bureaucracy (New Rochelle, N.Y.: Arlington House, 1969). The work was published originally in 1944.
  5. James Q. Wilson, Bureaucracy: What Government Agencies Do and Why They Do It (New York: Basic Books, 1989), p. 110.
  6. For an excellent series of initial recommendations in this area, see Scott A. Hodge, "A Budget Strategy to Reinvent the Federal Government," Heritage Foundation Backgrounder No. 1014, January 16, 1995. See also Ronald D. Utt, "Closing Unneeded and Obsolete Independent Government Agencies," Heritage Foundation Backgrounder No. 1015, January 25, 1995, and Scott A. Hodge, ed., Rolling Back Government: A Budget Plan to Rebuild America (Washington, D.C.: The Heritage Foundation, 1995).
  7. See Hodge, Rolling Back Government. See also Stuart M. Butler, Privatizing Federal Spending: A Strategy to Eliminate the Deficit (New York: Universe Books, 1985).
  8. See Devine, "How to Cut the Federal Bureaucracy."
  9. See Patrick Korten, "Why Congress Should Not Undermine the Presidential Power of Appointment," Heritage Foundation Backgrounder No. 1044, July 24, 1995
  10. For an excellent discussion of political managers in the federal bureaucracy, see Robert Rector and Michael Sanera, eds., Steering the Elephant: How Washington Works (New York: Universe Books, 1987).

About the Author