Has Government Been ‘Reinvented’?

Donald F. Kettl
Donald Kettl
Donald F. Kettl Donald F. Kettl is a Nonresident Senior Fellow at the Brookings Institution as well as Professor Emeritus and former Dean of the University of Maryland School of Public Policy

May 4, 2000

Let me thank the Subcommittee for allowing me this opportunity to discuss the status of the Clinton administration’s “reinventing government” initiatives. This is a critical issue—not only for assessing the administration’s accomplishments but also for charting the strategies and tactics of the next administration, whoever is president. For the last seven years, I have been leading an effort at the Brookings Institution to assess the results of the Clinton administration’s “reinventing government” initiative. At this stage, we can draw the following conclusions:

    Reinventing government is a work in progress. If this were a college course, this professor would grade “reinventing government” as a “B” (see Exhibit 1 for grades for individual items in the course). The Clinton administration unquestionably has invested substantial energy into the project. There have been real results. But there remains much work to be done and the effort is no where close to being completed.

      Substantial downsizing of the federal workforce has in fact occurred—but substantial issues remain. Federal civilian employment is now at 1.8 million, its lowest level since 1960. During the Clinton administration, it has dropped 19 percent. The reductions are unquestionably real. The underlying question is whether the downsizing has produced a right-sized, right-skilled workforce.

        There have been big improvements in customer service and procurement reform. “Reinventing government’s” efforts at improving customer service and procurement have unquestionably made it easier to deal with the federal government. Some agencies, like the IRS, continue to struggle at balancing customer service with their other functions. Even in these agencies, however, the improvements have been substantial.

          The “reinvention laboratories” demonstrate the potential of reinvention. More than 340 “reinvention laboratories” have experimented with management innovations. The accomplishments of many of these laboratories are truly impressive. What is most needed at this point is an effective strategy to learn the lessons the labs’ efforts teach.

            There has been uneven attention in the “reinvention” effort to resolving the problems identified in GAO’s “High-Risk Areas” and OMB’s “Priority Management Objectives.” The reinvention effort has had difficulty in penetrating and resolving some of the government’s most troublesome management problems. Many of these problems are rooted in basic management systems, like information and accounting, which have not been the focus of the “reinventing government” initiative. Solutions will require new partnerships between the executive and legislative branches.

            • Throughout the government, application of “reinvention” has been uneven. Some agencies, like the Federal Emergency Management Agency, are now dramatically different. Some departments, like the State and Commerce Departments, have seen less-dramatic change. Future progress hinges on making more progress across the federal government.

              No matter who is president in 2001, reinvention will have to continue, even if under a different name. Even a casual look around the world reveals that reinvention and its cousins have become universal. Governments everywhere face inescapable pressures to do more with less. Whoever is elected president in November will have no alternative but to continue reinvention-style efforts, no matter how they might be labeled. The first priority in these efforts must be building a new partnership with the Congress to make government efficiency, effectiveness, equity, and responsiveness a high priority.

              Let me discuss these issues in turn.

              Downsizing and Budget Savings

              There is no doubt that the federal civilian workforce is now substantially smaller than at the beginning of the Clinton administration. As Exhibit 2 shows, the workforce has shrunk by 19 percent, to 1.8 million workers. These reductions bring the federal civilian workforce to a level lower than any time since 1960.

              The reductions have fallen unevenly throughout the federal government, as Exhibit 3 shows. Civilian employment in the Department of Defense has shrunk by almost 30 percent, with reductions in Energy (25 percent) and Housing and Urban Development (23 percent) not far behind. On the other hand, three cabinet departments have grown: Commerce (especially to manage the census); State (to cope with international pressures); and Justice (to increase the number of guards at federal prisons).

              These reductions have raised several questions:

                Would the downsizing have happened anyway? Before the Clinton administration took office, defense spending was shrinking. Lower defense spending surely would have produced a smaller Pentagon workforce. In fact, nearly seven-tenths (68.9 percent) of the workforce reductions came from the Department of Defense. DOD’s large share of the workforce reductions is not surprising: federal civilian employees in DOD accounted for 44 percent of all civilian employees in 1993. That fact, coupled with reductions in defense spending, accounts for DOD’s disproportionate share of the workforce reductions. However, it is likely that the DOD reductions are larger than would have been the case without “reinventing government.” It is virtually certain that the substantial workforce reductions elsewhere in the federal government would not have occurred without “reinventing government.” Thus, “reinventing government” added significantly to the downsizing and cost-reduction effort.

                  If “reinventing government” downsized the bureaucracy, did it rightsize it as well? Even though the Clinton administration did downsize the bureaucracy, it did not plan which positions the government needed to loose and which skills it most needed to keep. There was an effort to reduce the number of middle-level managers. In general, however, the downsizing occurred as a result of individual employees’ responses to the buyout the government offered. There is little knowledge about the resulting skill mix of the federal workforce. There was little advance planning of what skill mix the federal government needs for the future. As a result, there is no assurance that the downsizing produced a rightsized government, with the skill set the government needs for the future.

                    How does downsizing affect the federal government’s pending retirement crisis? The federal Office of Personnel Management estimates that 32.1 percent of the federal workforce will be eligible for retirement by 2004. Other estimates suggest that an additional 19 percent of federal employees will be eligible for early retirement—and that up to 80 percent of the Senior Executive Service could retire (combining those eligible for both regular and early retirement). Cutting the workforce was one thing. We now face the staggering problem—and unprecedented opportunity—of designing and staffing the workforce of the future.

                    The downsizing accounts for about half ($54.8 billion) of the $111.8 billion in savings the Clinton administration claims from “reinventing government.” The downsizing reductions are clear and the estimates are reliable. Of the other half, most come from savings in procurement ($12.3 billion) and improvements in agency management. These savings are difficult to prove, since it is impossible to say with certainty what the government would have spent for these functions in the absence of reform. It might be possible to contest some of these estimates. However, in my judgment these estimates are certainly reasonable.

                    One question often raised is whether the “downsized” workers have been replaced by private contractors. Good numbers on government procurement are notoriously difficult to produce. However, the General Services Administration Government Procurement Data System reveals that, from fiscal year 1992 through fiscal year 1999, the federal government’s total procurement budget actually dropped slightly, by about $1 billion. The number of contract actions—contracts negotiated—decreased by 48 percent. The number of defense contracts shrank more quickly than the number of contracts elsewhere in the government (a 52 percent decrease in the number of defense contracts, versus a 42 percent decrease elsewhere). Total defense spending via contracts shrank slightly (by 1 percent), while contract spending elsewhere in the government grew slightly (by 1 percent). Thus, it is difficult to make the case that the downsized workers produced an increase in contracting.

                    In short, “reinventing government” has significantly downsized the bureaucracy and produced substantial cost savings. However, the effort has not attacked the critical workforce issues facing the federal government.

                    Customer Service and Procurement Reform

                    The Clinton administration launched a major customer-service initiative as part of “reinventing government.” Under an executive order to develop customer-service standards, all federal agencies had been forced to confront whom they were in business to serve. Administration officials now count more than 4,000 customer service standards for more than 570 government agencies and programs.

                    The customer service effort has been widespread and unprecedented. It has achieved substantial success. For example, in 1995 the Social Security Administration was judged as having the best toll-free customer service—better, in fact, than the nation’s leading private companies. A customer-satisfaction survey (compiled by Arthur Andersen, the University of Michigan Business School, and the American Society for Quality and released in December 1999) showed that federal agencies scored 68.6, compared with the private-sector rating of 72. Agencies that provided direct services tended to receive higher satisfaction scores than those that wrote rules, imposed fines, or enforced laws. The survey was rudimentary and raised methodological questions. Agencies surveyed, for example, could choose the customers and the services on which they were judged. For example, the survey gauged the attitudes of the parents of Head Start students, who might be expected to have high opinions of the program, but not first-grade teachers, who instruct Head Start’s graduates.

                    Some parts of the government have struggled mightily with the customer-service problem. For example, the Internal Revenue Service must balance its goal of making tax compliance easier with its other principal task of maximizing tax collections. However, the IRS has not yet succeeded in balancing these two objectives. In part, this is because the job itself is daunting. In part, this is because Congress has sent the IRS complex, changing, and often conflicting signals.

                    Moreover, the customer-service effort raises difficult questions:

                      Who is the customer? Many government programs have multiple customers. Conflicts (and conflicting expectations) often arise, especially between those who receive services and those who pay for them.

                      How can we draw the customer connection? For many programs, especially regulatory and enforcement programs, government agencies attempt to get citizens to do things they might not want to do. How can they promote “satisfaction” in such circumstances?

                      How can government avoid hypersensitivity to customer wishes? Government sometimes tries too hard to please citizens/customers. That can promote management inefficiency and drive spending up.

                      How can government balance top-down and bottom-up pressures? Customer service requires government to be built from the ground up to promote satisfaction for citizens. Accountability, especially to Congress, requires government to be built from the top down. Ensuring a balance between these competing goals is a daunting task for government managers.

                      How can government balance customer service with other, often conflicting objectives? Government, of course, does not exist only to make citizens happy by providing services. It also seeks other goals, like equity and equality. Balancing these often-competing goals can often prove difficult.

                      These big issues raise a deeper question. The “customer” model is powerful and, in fact, has generated great progress. However, because government is responsible for a wider array of objectives than private-sector companies, it must pursue more than just customer service. Indeed, some of “reinventing government” harshest critics have complained about just this issue in suggesting that the pursuit of customer service is dangerous. It is hard to argue with a substantial effort to make government friendlier and easier to deal with. The Clinton administration has made great progress on this score. However, we continue to face important challenges in finding the right balance between responsiveness and other important public objectives.

                      Along with customer service, the Clinton administration has also achieved substantial progress in procurement reform. With procurement cards, reengineering of the procurement process, and more flexible results-based procurement processes, “reinventing government” has substantially transformed the government’s procurement system. The administration claims $12 billion in savings from procurement reform. These figures are difficult to confirm, since it is impossible to know what procurement costs would have been in the absence of the reforms. However, the claimed savings are, in my opinion, eminently reasonable. The procurement and customer-service reforms surely rank as one of “reinventing government’s” most impressive accomplishments.

                      Reinvention Laboratories

                      “Reinventing government” licensed “reinvention laboratories” throughout the government. The reinvention labs were designed as places where managers could experiment with innovative ways of delivering services more cheaply and effectively. In some cases, the reinvention labs institutionalized reforms that were underway before President Clinton took office. In general, however, the reinvention lab program provided top-level support and encouragement for these efforts. It also encouraged the spread of the effort throughout the federal government. More than 340 reinvention labs have sprung up.

                      For example, managers in radiology departments at Veterans Affairs hospitals developed new electronic links that reduced the need for on-call radiologists. Postal workers in Newton, Massachusetts saved $50 million with a “movers guide” and “welcome kit” that improved service and reduced the Postal Service’s costs.

                      The reinvention labs are a clear success. They demonstrate how much energetic managers can accomplish, with the right support. Because the labs operate within the federal government’s overall management framework, they also demonstrate how many of the federal government’s management problems are self-inflicted. If there is a problem in the reinvention lab program, it is the difficulty of cumulating the lessons learned, so that they can guide the broader reform movement. The small size of the headquarters operation of “reinventing government,” housed in the National Partnership for Reinventing Government, has made it very difficult to conduct the extensive evaluations that the reinvention lab effort needs.

                      Strategic Focus: High-Risk Areas and Priority Management Objectives

                      For years, the General Accounting Office has identified “high-risk areas” in which management problems threaten to reduce program effectiveness and increase program costs. The “reinventing government” initiative has not dealt directly with the problems that GAO has identified. There have been important initiatives, ranging from improvements in information technology and agency-specific reforms, like those in the IRS, General Services Administration, and the Federal Emergency Management Agency. Reinvention, however, has focused more on downsizing and operational improvements than on the broad, systemic management problems that GAO has identified in its high-risk list.

                      The Office of Management and Budget has identified “priority management objectives,” ranging from managing the Year 2000 computer problem and acquisition reform to improving Department of Energy contract management and modernizing student aid. These objectives are fully consistent with “reinventing government” and, in fact, represent cooperative efforts between OMB and the agencies involved. The agencies, for their part, have launched their own reforms in the spirit of “reinventing government.” However, efforts to date suggest two conclusions. First, these “priority management objectives” have depended on OMB’s energetic pursuit of the goals. They have been less important to the “reinventing” effort, except to the degree that “reinventing government has encouraged agencies to develop their own aggressive reforms. Second, significant problems remain in tackling many of these “priority management objectives.” The Y2K effort was surely an unqualified success. But from IRS reforms to managing risks in the space program, major challenges remain.

                      The “reinventing government” initiative has developed its own list of “high-impact agencies.” The goal is to identify those agencies that produce programs that citizens care most about—and to focus the agencies’ attention on delivering better results. The endeavor is at an early stage but shows important signs of progress. Some of the efforts reinforce the focus of GAO and OMB on more systemic management problems. However, because “reinventing government” has tended to concentrate on downsizing and process improvements (like customer service and procurement reform), it has had less impact on the problems that OMB and GAO have identified.

                      It is clear, moreover, that reinvention has been an uneven phenomenon throughout the federal government. A 1996 Merit Systems Protection Board survey, more than three years after the NPR’s launch, showed that only 37 percent of federal employees believed that their organization had made reinvention a top priority. The NPR’s management improvement goals penetrated far less into the Pentagon than in civilian agencies. Morale in many agencies was poor. Only 20 percent of federal workers said that the NPR had brought positive change to government. Where the NPR was a top priority, 59 percent of employees thought productivity had improved; where it was not, only 32 percent thought productivity was better. Where the NPR’s goals received emphasis, employees were three times as likely to think that government organizations had made good use of their abilities. They were almost twice as likely to believe that they had been given greater flexibility. Government employees’ attitudes varied with how much top managers made reinvention a priority.

                      Next steps in “reinventing government” will have to deal more directly with the fundamental structural problems that GAO and OMB have identified. It will also have to ensure wider buy-in across the federal government.

                      The Future of “Reinvention”

                      While the Clinton administration’s reinventing government initiative unquestionably encountered serious problems, it nevertheless produced genuine accomplishment in its first five years. It saved a significant amount of money, brought substantial managerial reforms (especially in customer service and procurement processes), and promoted a more performance-based discussion on government’s functions.

                      “Reinventing government’s” shortcomings, though, are as instructive as its early successes. President Clinton’s bold proclamation about the end of “big government” missed the far more important, if much more subtle, transformation in the way government works in the United States. The federal government is no longer organized for the job that law and the Constitution charge it to do. Devolution, contracting out, and other third-party strategies have grown significantly. The federal government’s capacity to manage these programs has not grown to keep pace. Many of government’s most significant management problems are a direct result of this mismatch of capacity and strategy. Without fundamental reforms in civil service, information technology, and financial management, this mismatch will continue to grow. The result can only be more tales of a government poorly equipped to accomplish its most important work.

                      With the budget deficit disappearing, the defining reality of American domestic policy since the late 1960s has evaporated. Unlike the fall of the Berlin Wall—the defining reality of a generation of American foreign policy—the demise of the budget deficit scarcely brought dancing and celebration. That is because two other deficits, quieter but just as important, remain: government’s performance deficit—the gap between government’s goals and its results—and its confidence deficit—citizens’ trust in government’s ability to get its job done. Management reform—improving government’s ability to achieve its results—is the key to reducing the performance deficit. Reducing government’s performance deficit, in turn, is the key to progress on the confidence deficit.

                      Making government work better thus is a goal that both political parties have little choice but to embrace. Indeed, management reform has become a truly global phenomenon, and American elected officials, of both parties, have little choice but to continue to pursue such reforms in the future. The public is unlikely to reward candidates for improving the way government works—Americans naturally tend to expect their public services to work smoothly and have often balked at rewarding government workers for doing their tax-funded jobs. But government officials also have little alternative but to work hard to improve performance, if only because embarrassing failures can have electoral repercussions—and because every penny of increased productivity is a tax that does not have to be raised or an expenditure that does not have to be cut. That is what makes a continued effort to reinvent American government an inevitable if, perhaps, thankless task. A constant battle for management reform is the one sure bet in American politics, regardless of who is elected president in November.


                      Back to

                      Exhibit 1

                      The “Reinventing Government” Report Card






                      Accomplished the goal, but planning to match the downsized workforce with agency missions was weak.

                      Identifying objectives of government


                      The NPR sought in 1995 to focus on what government should do—but the effort evaporated as the Republican threat faded.

                      Procurement reform


                      Fundamental transformation of procurement system. Some vendors complain, but the system is far more efficient than it was.

                      Customer service


                      Great progress in some agencies, but major failures in othersnotably the IRS.

                      Disaster avoidance


                      Substantial efforts in many agencies, notably FEMA. Spectacular failures in others, notably the IRS. The big test: the Y2K problem.

                      Political leadership


                      Consistently strong leadership from the top but inconsistent below. Federal workers have gotten mixed signals.

                      Performance improvements


                      Linkage of NPR with the Government Performance and Results Act is inconsistent.

                      Improved results in “high-impact” programs


                      Good strategy—but likelihood of achieving goals is low.

                      Service coordination


                      Efforts to improve the coordination of service delivery are embryonic.

                      Relations with Congress


                      Efforts to develop legislative support for NPR initiatives have, with the exception of procurement reform, been weak and ineffective. Support from Congress: poor.

                      Improvements in citizen confidence in government


                      The steady slide in public trust and confidence in government has ended, but that has more to do with a healthy economy than improved government performance.

                      Inspiration from other governments, private-sector reforms


                      Wide survey of other ideas—but more a grab-bag of options than a careful analysis of which ones fit federal problems.



                      No administration in history has invested such sustained, high-level attention to management reform efforts.

                      OVERALL GRADE



                      Substantial progress made over first five years, but much more work lies ahead. Successive administrations will have little choice but to continue the NPR in some form.

                      Source for report card:

                      Donald F. Kettl, Reinventing Government: A Fifth-Year Report
                      (Washington: Brookings Institution,

                      Back to

                      SOURCE: U.S. Office of Personnel Management
                      Office of Workforce Information
                      Monthly Report of Federal Civilian Employment (SF 113-A)
                      Revised March 28, 2000

                      Back to text

                      Exhibit 3

                      Federal Executive Branch Employment
                      EXECUTIVE BRANCH (Excludes USPS & Postal Rate Commission) -19.0
                      Defense   -29.6
                      Energy   -24.7
                      Housing and Urban Development   -23.1
                      Agriculture   -18.0
                      Veterans Affairs   -16.5
                      Interior   -14.6
                      Labor   -10.2
                      Transportation   -9.3
                      Treasury   -7.4
                      Education   -5.0
                      Health and Human Services (includes Social Security Administration)   -4.4
                      Commerce*   16.6
                      State   19.6
                      Justice   27.8
                      Federal Deposit Insurance Corporation   -67.1
                      General Services Administration   -31.8
                      NASA   -26.7
                      Environmental Protection Agency   -0.6
                      Federal Emergency Management Agency   9.2
                      * Does not include 31,587 temporary census enumerators.    
                      Figures are preliminary and subject to agency revisions.    
                      SOURCE: U.S. Office of Personnel Management    
                            Office of Workforce Information    
                            Monthly Report of Federal Civilian Employment (SF 113-A)    
                            March 25, 2000