Reviewing the National Performance Review

George Nesterczuk is majority staff director for the Civil Service Subcommittee of the House Committee on Government Reform and Oversight.

Five months after the 1994 election, President Clinton launched an ambitious project to deliver on his campaign promise to “radically change the way government operates—to shift from top down bureaucracy to entrepreneurial government.” Vice President Gore, who was given the task of “creating a government that works better and costs less,” quickly assembled a team of career government executives and outside management consultants and embarked on the National Performance Review (NPR). By September 1993 they delivered the blueprint that was to serve as the Clinton administration's reform agenda (REGO I). The reinvention, which was meant to evolve over time, was itself reinvented in the wake of the Republican victory in the 1994 congressional elections. Chastened by the electorate, President Clinton proclaimed that he had received the voters' message “loud and clear” and proceeded to launch Phase II of reinvention (REGO II). The contrast between the two reform agendas could not be more striking and the change in course, including the reshaping of the coalition to promote REGO II, is a testament to the adaptability of the Clinton administration.

In this century alone there have been eleven major initiatives undertaken to reform the administrative practices of the federal government. One of the most recent was issued by the Grace Commission, which delivered its recommendations during the first term of the Reagan administration. The erosion of public confidence in the institutions of government has made reforming government a durable political issue for a generation of politicians of every persuasion. For the past thirty years, opinion polls have tracked the growing public disenchantment with government. Citizens view government as too large, too expensive, too intrusive, wasteful, and inefficient; however, not all agree on which specific services and functions should be cut back or eliminated.

Of course for the ideologically motivated the choices are clear. For those on the right who believe that government is best which governs least, no program is sacrosanct. For them reform is synonymous with the termination of programs and agencies, the privatization of functions and services, the sale of assets, and the devolution of federal responsibilities to state and local governments.

Supporters of an activist government, on the other hand, can always find a redeeming feature in every program. They tolerate reforms that scale back spending on certain programs as long as the savings are reinvested in other programs.

Those not burdened with ideological baggage rely on their personal values to judge the merits of government, and tend to balance self-interest with a sense of fairness to others. This naturally results in a wide range of opinions on the relative worth and importance of government programs.

Despite this wide range of views as to what might constitute appropriate government reform, there are some aspects of reform that appeal to all. A consensus can always be formed in support of reforms that eliminate fraud, waste, and abuse, and reforms that increase efficiency and productivity. A reform agenda that contains these elements permits broad-based support and helps mute any opposition. Yet the elimination of fraud, waste, and abuse in government remains the elusive apple pie of contemporary politics.

In launching the NPR, the Clinton administration had a clear vision of the coalition it would need to successfully promote its agenda. The coalition shaped and focused the agenda by eliminating issues that constituent groups found unacceptable. The stakeholders included the bureaucracy (particularly federal employee unions), public policy experts, public administration advocates, environmental groups, health care reformers, consumer advocates, veterans, seniors, business executives, and other representatives of President Clinton's core constituencies. Drawing on input from these constituents and various government agencies, the National Performance Review report was drafted in September 1993. The focus on process, not function, is unequivocally stated in the report: “The National Performance Review focused primarily on how government should work, not on what it should do. Our job was to improve performance in areas where policymakers had already decided government should play a role.”

In piecing together the reinventing government agenda, the administration borrowed liberally from past reform efforts. The Grace Commission's recommendations of financial management, debt collection, and procurement reform were resurrected, as were recommendations for increased use of information technology, consolidation of duplicative functions, and elimination of bureaucratic layering. Management coordination councils and interagency groups in the executive branch generated a steady stream of recommendations on improving management systems.

Improvements can and will always be made. But the synthesis and packaging needed to effect change is unique to each administration. The Grace Commission, for example, highlighted the benefits of competition in improving efficiency and controlling costs. Privatization and outsourcing were the favored means to establish benchmarks for efficiency. Private-sector loan portfolio management, debt collection, and cash-flow management techniques were implemented. Savings from tougher enforcement of fraud, waste, and abuse oversight through the inspectors general (IGs) were also emphasized.

The Clinton administration also borrowed concepts from the private sector. Total quality management (TQM), management by objectives (MBO), and business process re-engineering (BPR) principles were modified in an attempt to fit the government environment where no competition exists and there is no bottom line by which to measure success or failure. The administration sought to synthesize these concepts into a set of principles that would create an entrepreneurial government. In the words of President Clinton: “Our goal is 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.”

The desire to move the bureaucracy away from a culture of complacency and entitlement is not new. Both the Carter and Reagan administrations enacted major personnel management initiatives such as promoting better pay for better performance, using incentive awards, and removing workers who perform poorly. But the notion of empowering frontline bureaucrats with greater autonomy is definitely unique to the Clinton administration. In order to achieve this cultural change the administration settled on four guiding principles.

The reinvention agenda is overwhelmingly focused on process not function; the “how” of government, not the “what.” And the processes targeted for reform are almost exclusively internal to government. On this point the NPR is very specific. During the campaign Clinton promised to eliminate one hundred thousand federal jobs. As he prepared his first budget proposal, which specifically called for deficit reduction, he needed to generate spending reductions to make the proposed tax increases more defensible.

The NPR provided a useful source for these reductions by generating savings from relatively painless systemic reforms rather than from politically difficult program cuts. The administration targeted the government's internal administrative positions as appropriate sources for personnel reductions. They included supervisors, headquarters' staffs, personnel specialists, budget analysts, procurement specialists, accountants, and auditors. Additional reductions would come from re-engineering work processes to achieve higher productivity, thereby eliminating unnecessary layers of management. Managers' span of control was to be increased from a ratio of one supervisor for every seven employees (the current government-wide average) to one supervisor for every fifteen employees.

A scan of the Central Personnel Data File produced a count of just over 500,000 federal employees (out of a total work force of 2,100,000) engaged in these “central control” activities. Using private-sector estimates that these overhead jobs could be cut in half, the administration gave the president a reduction target of 252,000 positions, with a projected five-year budget savings of over $40 billion. As long as the primary focus was on eliminating supervisory and management positions, employee unions were on board with the agenda to eliminate federal jobs.

Other key constituencies also needed to be accommodated. Among the “central control” activities are a vast number of equal employment opportunity specialists, employee relations specialists, and a substantial number of female employees; however, one of reinvention's personnel goals was to promote diversity in the workplace. Thus reductions had to be steered away from preferred groups. The review proposed the use of buyouts to permit more targeted reductions, and Congress provided the vehicle with the passage of the Federal Workforce Restructuring Act of 1994.

The act slated for termination an additional 20,900 full-time equivalent (FTE) positions and established a five-year schedule with annual ceilings for all 272,900 positions. The act authorized the use of voluntary separation incentive payments of up to $25,000 per employee to encourage resignations and retirements. The administration now had the tools it needed to begin reshaping the work force.

The second goal of putting customers first was more difficult than meeting the targeted personnel reductions. In the days when the government was populated with civil or public servants rather than government employees, the notion of customer service was part of the culture. The sense of public service in government employment has long since disappeared and has been replaced with complacency and a belief in job entitlement. Thus the lack of a clear definition of “customer” can lead to unintended consequences.

The NPR stipulated very clearly, “By `customer,' we do not mean `citizen.'” Customers are those who receive a direct benefit from the services rendered. This definition easily permits catering to special interests and clients of the welfare state. But satisfying the needs of a welfare constituency or of the National Education Association may be at odds with the best interests of the taxpaying public. Unfortunately these ambiguities and inherent conflicts of interest are woven into the reinvention agenda.

The third guiding principle of empowering employees also contains some contradictions. The strategy calls for replacing traditional command-and-control management with decentralized decisionmaking balanced by greater accountability for results. Partnership councils between management and labor were created to define the expectations and measures against which employees would be held accountable. This rhetoric flies in the face of actual implementation. A number of recommendations in the agenda address the need for improved performance and increased accountability, but all the systems by which these actions could be measured or enforced have been weakened. The audit functions and the IGs were specifically targeted for downsizing. Reports to Congress were targeted for elimination. Management and supervisory positions were scaled back, and the performance management system in force at the time was allowed to expire. In its place the administration has aggressively promoted a pass/fail system for employee performance evaluation.

The NPR included a proposal to streamline and consolidate the civil service appeals system that severely restricts managers' ability to deal with poor performers. That proposal was withdrawn in the face of opposition from federal unions and civil rights groups. The employee empowerment element of the reinvention calls for radical decentralization of decisionmaking power while simultaneously gutting effective oversight and abolishing the means to evaluate individual performance. The administration purposefully chose to move in the direction of group performance and organizational accountability.

Of even greater concern was the clear intent to transfer power from Congress to the bureaucracy, and to lower levels of decisionmaking within the bureaucracy. The review included proposals to move to multiyear appropriations, pass a presidential line-item veto, eliminate many reporting requirements, and redefine the role of the IGs from congressional watchdogs to management consultants. As part of the NPR human resources reforms, the administration submitted a legislative proposal that would have allowed every government agency to establish independent personnel systems without congressional intervention. This would have dismantled the civil service system and replaced it with personnel systems in which every aspect of employment would be subject to labor bargaining. The assurance that the public's interest would be protected at the bargaining table was to be provided by the requirement that negotiations be constrained by the need to bargain to “a good government standard.” What such a standard might entail is defined nowhere.

The final guiding principle, which called for cutting back to basics, provided the second source of significant budget savings. The NPR called for reorganizing agencies, consolidating duplicative functions, closing and consolidating regional and field offices, and consolidating various other programs, notably training for the unemployed, food safety inspections, and education grants. Recommendations for specific program terminations included wool and mohair subsidies, price supports for honey, unobligated funds for highway demonstration projects, and reduced Essential Air Service subsidies. The review also called for increased cooperation with state and local governments to garner savings through more effective local administrating of federal programs.

Table 1 Estimated Savings from NPR Recommendations
FY95-FY99 in billions of dollars	1993 estimate	1995 update
Personal reductions	$ 40.4	$ 40.4
Procurement reforms	$ 22.5	$ 12.3
Information technology	$ 5.4	$ 2.3
Reducing intergovernmental administrative costs	$ 3.3	$ 2.8
Changes in individual agencies	$ 36.4	$ 2.8
Total savings for REGO I	$ 108.0	$ 57.8
Source: Common Sense Government Works Better & Costs Less. Third report of the National Performance Review, September 1995.

In some of his previous budget submissions, Clinton used the NPR's projected budget savings to offset increases in program spending. The estimated savings from REGO I amounted to a total of $108 billion from FY95-FY99. Table 1 identifies these projected savings by category.

Two years after the original estimates of REGO I savings were made, the administration is unable to sustain many of the original assumptions. Expected savings from procurement reforms have been cut in half. Much of the projected savings from the increased use of information technology have come from enhancing productivity, which has resulted in staff reductions. The productivity savings already are reflected in the entry for personnel reductions, thus the technology estimate also has been reduced.

Savings attributable to actions in individual agencies are difficult to isolate from other related actions and are proving hard to quantify. As of September 1995 the administration could verify only $2.8 billion in savings from actions already taken. Undoubtedly this number will grow, but many of the proposed actions result in “point in time” savings estimates. As circumstances and statutes change one or two years after such estimates have been made, it becomes difficult to replicate the original assumptions. These estimated savings figures are very soft. Neither the General Accounting Office (GAO) nor the Congressional Budget Office (CBO) could verify or track most of these assumed savings.

The most easily quantified savings are the personnel reductions. Payroll and program expenditures are the most tangible indicators of government size and activity. Agency reorganizations and consolidations, while useful and sometimes important, seldom generate budgetary savings unless they result in staff reductions. Likewise process reforms can improve the quality of services, but unless they translate into increased productivity with attendant staff reductions, they do not contribute to budget savings. For these reasons reform strategies often rely on mandated staff cuts to force bureaucrats to impose process changes that will improve productivity.

To reduce the size and scope of government and to reduce spending, the best approach is to terminate or scale back programs and functions. The remaining functions then can be staffed with the right skills and number of employees. This approach to reinvention was not a significant part of the strategy in Phase I of REGO.

Through FY95 personnel reductions totaled 166,000 full-time equivalent positions below the level inherited by the administration. (FTE is defined as one work year or 2080 staff hours.) A March 31, 1996 staffing report shows that reductions through the end of FY96 should total approximately 200,000 FTE less than FY92 base levels.

The administration sometimes uses a base for FTE from FY93 budget estimates. Estimated budget figures are projections of future expectations. When actual FTE numbers are available, there is no reason to use projected estimates. For this reason, this analysis uses actual FTE levels for previous fiscal years and projected estimates for current and future fiscal years. The base year against which to measure the impact of the administration's reinvention therefore is FY92. Table 2 provides federal civilian employment statistics in defense and nondefense agencies for the period FY81-FY97.

As Table 2 indicates, of the 149,000 FTE reductions from FY92-FY96, fully 75 percent are the result of reductions in the Department of Defense (DOD). They represent a 16 percent drop from the 1992 base level and have little to do with reinvention. The DOD downsizing is the “peace dividend” that came with the end of the Cold War. Defense reductions began in 1989, primarily driven by the Base Realignment and Closure Commission. As the result of a reassessment of DOD's mission, significant staff reductions and reorganizations have taken place and will continue until approximately FY2002.

Table 2 Federal Employment in the Executive Brance by FTE in Thousands
	FY81 (est.)	FY85 (est.)	FY92	FY93 (%)	FY94 (%)	FY95	FY96	FY97	81-85	81-85	92-96	92-96
Defense	916	1008	949	932	868	822	800	767	+92	+10%	-149	-16%
Nondefense	1164	1085	1182	1202	1179	1143	1135	1137	-79	-7%	-47	-4%
Total	2080	2093	2131	2133	2047	1964	1935	1905	+13	+0.6%	-196	-9%
Source: OPM Employment Trend reports and the OMB budget documents.

By the end of FY96 nondefense reductions will be approximately 50,000 FTE, a 4.2 percent drop from FY92. This represents only 25 percent of all the reductions achieved since 1992 despite the fact that nondefense employees account for 55 percent of executive branch employment. Clearly, defense reductions represent a disproportionate share of the administration's claimed reinvention achievements. It is noteworthy that in the first year of the Clinton administration nondefense employment actually rose by 20,000 FTE. The use of FY93 rather than FY92 as the base period against which to count the administration's employment reductions distorts the true impact of the administration's policies on federal employment.

During the comparable period in the Reagan administration, nondefense reductions totaled 78,600 FTE, or 7.2 percent of 1,163,600 FTE. The FTE levels for FY80 are not available. However, since the Reagan administration imposed a hiring freeze on the bureaucracy in January 1981, it would be correct to conclude that FTE usage in FY80 would have been somewhat higher than the FY81 figure in Table 2. Using FY81 as a base rather than FY80 understates the magnitude of the cuts achieved by the Reagan administration in nondefense employment. The Reagan reductions also must be viewed in the context of working with a hostile Congress that sought to prevent any staff reductions. In some instances Congress resorted to imposing employment floors in some of the agencies in order to prohibit additional reductions.

As indicated in Table 2, average reductions in nondefense agencies total 4 percent from FY92-FY96. However, experience at individual agencies varied widely from an expected 16 percent increase at the Department of Justice to a 42 percent decrease at the Federal Deposit Insurance Corporation. In addition the EPA, DOE, FEMA, and SBA all posted some growth. A number of agencies show some decline in staff levels (reductions of 7 percent or less). The noteworthy standouts where reinvention seems to have had an impact include OPM (-32 percent), HUD (-15), AID (-19), GSA (-19), NASA (-11), and USIA (-11). The amount of reductions definitely accelerated beginning in FY95. From FY96, however, it becomes increasingly difficult to differentiate the impact of REGO I from other congressional appropriation and budgetary actions.

As a matter of interest, nondefense reductions during the Reagan years were more uniform. The average reduction was 7 percent across the agencies with the following exceptions: Justice (+8 percent), State (+6), USIA (+6), VA (+4), Treasury and NASA (0). Agencies taking much larger reductions included: Education (-25), GSA (-22), HUD (-21), OPM (-20), and Labor (-15). All other agencies achieved their targeted reductions ranging from 8 to 10 percent. The Reagan strategy mandated staffing reduction targets to force process changes in the bureaucracy. This was a different approach than the REGO I strategy, and apparently more effective. The Reagan priorities called for strengthening the defense and foreign policy establishments. It is interesting that HUD appears on both the Reagan and Clinton lists for significant reductions, as do the two central management agencies OPM and GSA. OPM today has less than 4,000 FTE, down from 7,200 in FY81. GSA has about 16,000, down from 33,000 in FY81.

Finally, it is unclear to what extent the administration has engaged in “phasing over” rather than “phasing out” some of its workload. Since programs are not being terminated except where Congress has intervened, the work previously performed by past-employees may still need to be performed. This can be achieved by outsourcing to government contractors, and some of the reinvented work already has shifted to contractors, but no one knows precisely how much. The DOD certainly relies heavily on contractors to leverage its work and of the 150,000 FTE reductions the agency has taken, perhaps as much as one-third has been privatized to contractors.

At least two reinvention efforts have been outsourced. One was OPM's training activities which were turned over to a nonprofit contractor. This resulted in a reduction of 250 FTE. Another was outsourced by creating an ESOP for 700 OPM employees who took over the business of performing background investigations for agencies on federal employees. While the practice of outsourcing compensates for staff reductions, it does not generate the same budget savings as reducing the actual workload. That is because it substitutes a contract expenditure for the payroll expense to employees. Contract labor is often cheaper than government employment, but the savings tend to be in the range of 20 to 30 percent of an FTE.

Reductions in management and control positions (personnel, budget, accounting, procurement, and audit) were central to the REGO I aim to achieve staff reductions of 272,900 FTE by FY99. In fact the NPR reduction targets were so specific that the administration went to Congress for authority to use cash incentives to induce employees to leave voluntarily. In the Federal Workforce Restructuring Act of 1994, Congress authorized the use of these voluntary separation incentives for up to $25,000 per employee. The authority was provided for a limited time in FY94 and FY95. The rationale was that this would give the administration the flexibility to target reductions along reinvention lines without resorting to more cumbersome reduction in force (RIF) procedures. The law required that there be one FTE reduction for each buyout taken.

The OMB assumed responsibility of managing the buyout process and interpreted the “one buyout to one FTE” requirement loosely. As a result the buyouts turned into a golden handshake that cost the taxpayers $750 million in nondefense agencies alone. Several agencies offered buyouts even though staffing levels increased during the buyout period. The Department of Justice is noteworthy in this respect because it is the only agency that has been hiring steadily since 1993. The department nevertheless bought out 678 employees in FY94 and 187 in FY95. There are similar instances of questionable buyouts in other agencies, as well as examples of agencies where buyouts exceeded staff reductions.

According to a recent GAO report, 72 percent of buyouts went to employees eligible for retirement. A large portion of these employees would have retired without the $25,000 incentive. Normal retirement rates dropped significantly in the first year of the buyouts as employees waited for their separation incentives. The cost of buyouts for fiscal years 1993, 1994, and 1995 now stands at $2.8 billion.

What about the original stated purpose for the buyout authority—to reduce staff levels in control positions? The GAO report concludes that the administration is way off target, although there is still time for improvement before 1999. More buyouts went to nontargeted occupations (60 percent) than to employees in control positions. In an audit of twenty-four agencies, the GAO found that buyouts accounted for 48 percent of agency staff reductions in FY94 and FY95. RIFs generated 6 percent of the reductions while normal attrition accounted for the rest.

The number of employees in management control positions was reduced, but as a proportion of the remaining work force, little has changed. The original goal was to reduce the ratio of supervisory positions by 50 percent (from 1:7 to 1:15). The result to date is as follows: personnel positions down 8%, budget positions down 1%, audit positions down 3%, procurement down 12% in defense and 2% in nondefense agencies, and headquarters staff down 2% in defense and 6% in nondefense agencies. Overall, the supervisor ratio is now 1:7.5. The 50 percent reduction target for these positions was to be achieved by 1999, but it will take a drastic change in strategy for the administration to achieve this goal on time. There are indications that the bureaucracy has stopped trying to achieve these particular targets.

The GAO also confirmed that agencies have resorted to converting large numbers of their supervisors into “team leaders” to reduce the number of “supervisory” positions. For example, the Bureau of Land Management achieved 40 percent of their supervisor reductions through conversions. The comparable figure for the Federal Aviation Administration is 35 percent and the Department of Health and Human Services, 28 percent. Conversions distort not only the data on supervisory ratios, but also the total FTE reduction data since conversions do not result in real staff reductions.

The original NPR agenda as embodied in REGO I contained recommendations covering twenty-seven agencies and fourteen government management systems. REGO I consisted of 254 major recommendations affecting agencies and 130 recommendations to reform the management systems. A total of 1,203 specific action items were promulgated to implement the recommendations. In September 1995 Congress asked the GAO to review the status of these action items. In an April 1996 report the GAO verified that 293 were complete and 67 had been partially implemented. No attempt was made to perform a qualitative assessment or verify estimates of budget savings from the items. The status of the remaining items is currently unknown. However, many items may have been overtaken in priority by the unveiling of REGO II.

Following the midterm elections in 1994, the administration launched a new reform initiative as Phase II of the reinvention, or REGO II. The new agenda was much more in tune with that of the newly elected Republican majority in Congress. In the words of President Clinton, “We know we have to go beyond cutting, even beyond restructuring, to completely reevaluate what the federal government is doing. Are we doing it well? Should we be doing it at all? Should someone else be doing it?” This is a far cry from the charge in REGO I to look at “how” the government is doing its business and avoid the issue of “what” the government is doing. In a January 1995 memo to agency heads, Vice President Gore spelled out the “second phase” of the NPR as follows:

[Moreover] Phase II of NPR will also review the federal regulatory process to find a way to get better results for the public with less interference in their lives.

The four basic principles of REGO II promote the termination of programs and agencies, the privatization of appropriate functions, the devolution of funds and responsibilities to state and local governments, and consolidation and streamlining to further shrink the work force. This reform agenda is much more threatening to the status quo than REGO I.

REGO II proposals include more than 186 new recommendations for agency actions and 121 regulatory reforms. Unfortunately as Congress proposed to enact its version of REGO II through appropriations bills and various balanced budget acts during the latter half of 1995, it found the administration less enthusiastic about serious reforms. In exercising the veto over specific reduction measures, Clinton seriously undermined the credibility of his own reform agenda.

The credibility of REGO I and REGO II was further brought into question with the president's FY97 budget submission. In that budget staff levels for the executive branch are slated to decrease in defense by 32,000 FTE but show a net increase of 2,000 FTE for nondefense agencies. In testimony before the Senate, Veterans' Affairs Secretary Jesse Brown responded that the president assured him that the VA budget was not binding. Brown said, “I am not planning to live with it. I am not planning to live with your budget . . . nor am I planning to live with the president's line.” NASA administrator Dan Goldin took the same tack, “The White House has instructed us to take no precipitous action on out-year budgets, and we are taking them at their word.”

The administration continues to claim that it has “proposed eliminating more than 400 obsolete programs and is in the process of closing more than 2,000 unnecessary field offices.” In testimony before a House panel, however, on two separate occasions the OMB witnesses could not identify a single terminated program nor specify where field offices had been closed. On closer examination the administration proposed to terminate only 130 programs—the remaining 270 were to be consolidated into 27 giant programs.

Congress on the other hand seems to have taken control of the reinvention reins. The House Appropriations Committee reports that Congress has terminated over 270 federal programs at a savings of over $4 billion. Total discretionary spending was reduced in FY96 by $16 billion from FY95 levels. And more serious reforms are under way in the areas of welfare and health care. Clinton appears to be giving up the opportunity to lead his own reform agenda. Either it has proven too unpopular with his core constituencies or he may no longer need to use reform as a political issue.

Clinton is fond of proclaiming that he has taken the country to the smallest government since the days of John F. Kennedy. A closer look at the facts makes this claim somewhat problematic. In 1961 federal employment stood at 1,950,000; today it stands at about 1,950,000. In Kennedy's day the Department of Defense employed 1,150,000 people, today that number is 800,000. In Kennedy's day the nondefense agencies counted 800,000 employees, today that number is approximately 1,150,000. Considering that the public's disenchantment with big government extends overwhelmingly to nondefense functions, Clinton's nostalgic references to the Kennedy years provide a useful target for the future. Reducing nondefense employment from the current level to the 800,000 level of Kennedy's day would be a challenging but eminently commendable goal.

Regulation is published four times a year by the Cato Institute. Editorial and business offices are located at 1000 Massachusetts Avenue, N.W., Washington, D.C., 20001. For subscription information, please write to Circulation Department, Cato Institute, same address, or call (202) 842-0200. Send email inquiries to subscriptions@cato.org, or subscribe online via the World Wide Web at: http://www.cato.org/pubs/regulation/reg-ordr.html