25 Years Later, What Happened to 'Reinventing Government'?

The ambitious public management crusade of the 1990s has made a mark on governments everywhere. But it’s fallen short of some of its goals.
by | September 2016
(Illustrations by Trevor Andersen)

In the fall of 2011, Washington Gov. Christine Gregoire convened a meeting to review one of the proudest accomplishments of her two terms in office -- the state’s Government Management Accountability and Performance system. GMAP was inspired by earlier experiments that had succeeded in the New York City Police Department and in Baltimore. Gregoire had worked for six years to put a similar system in place in Washington. Hundreds of performance measures had been identified and linked to strategic agency plans. The governor and her leadership team had held quarterly meetings -- all open to the public -- with the heads of 30 agencies to work on specific problems in eight priority areas (health care, public safety, transportation, vulnerable children and adults, education, natural resources, economic vitality, and government reform and efficiency).

GMAP helped agencies such as Child Protective Services bring a new focus to their mission. Following up quickly on referrals about potential child abuse can prevent serious, even fatal, injuries. When GMAP was launched in February 2005, only 69 percent of referrals were acted on within 24 hours. By focusing attention on this metric, GMAP pushed the agency to streamline its processes and improve response times. In two years, the quick response number rose to 90 percent. Such results, achieved by departments all across the executive branch, won widespread praise. In 2007, the Council of State Governments gave the state its first governance transformation award. By “measuring results and delivering practical, useful tools and solutions,” the council found, “the program is driving accountability and helping make Washington state government better.”

As Gregoire approached the end of her second term, she decided to take stock of the system. She asked Wendy Korthuis-Smith, GMAP’s director, to assemble a group of agency directors to discuss what was working, what wasn’t working and what changes could make GMAP more effective. Gregoire and Korthuis-Smith envisioned a tune-up. What they got was a far more serious warning.

“Some of the directors felt like the measures weren’t relevant anymore but that they couldn’t or shouldn’t change them,” says Korthuis-Smith. Some saw the GMAP meetings with the governor as “dog and pony” shows. “The system we’d built didn’t feel flexible and nimble,” she says. “There was too much focus on polishing presentations, when what we really wanted were cycles of experimentation, improvement and results.” As the conversation proceeded, Korthuis-Smith realized that unless major changes were made, GMAP ran the risk of becoming just another stale compliance regime.

Other places that have tried ambitious performance management systems have encountered similar complaints. Of course, it’s not surprising that stubborn managers would resist accountability. What is surprising is who’s doing the resisting -- some of the most innovative people working in government today.

Take Mike Flowers, chief analytics officer during the last years of the Bloomberg administration in New York City. Flowers was a leading figure in the effort to use civic data to solve city problems. Yet his assessment of the city’s performance reporting system is decidedly mixed. The accountability component -- “sending a message that you don’t have a right to be incompetent” -- is good, Flowers says. But city agencies get focused on pumping up their numbers. Instead of paying attention to new problems or looking for new methods of problem-solving, managers focus on what they already do well. When Flowers first met the head of Maryland’s acclaimed StateStat performance measurement system at a conference, he did not hide his belief that the approach did more to inhibit problem-solving than to promote it. “I was very honest,” says Flowers. “I told her, ‘You people just make matters worse.’”

What Flowers found was that even the best performance measurement systems tend to ossify. In 2010, 11 state and local public interest associations joined together to form the National Performance Management Advisory Commission. In its report, A Performance Management Framework for State and Local Government, the commission singled out Nashville, Tenn.’s Results Matter system as an example of a robust success in managing for results. But when Kristine LaLonde became Nashville’s co-chief innovation officer two years after Results Matter was cited as a national model, she found a city government suffering from what she calls “Results Matter PTSD.”

“The department heads I have the most respect for hated it the most,” says LaLonde. Skeptics saw it as a system created by outside consultants with no knowledge of agencies’ true challenges. Instead of promoting problem-solving, it encouraged agencies to post good numbers. Suspicions about the accuracy and meaningfulness of those numbers were widespread. “Many people didn’t trust the system,” LaLonde says.

That wasn’t how it was supposed to work. A generation ago, governments across the United States embarked on ambitious efforts to use performance measures to “reinvent” how government worked. Much of the inspiration for this effort came from the bestselling 1992 book Reinventing Government: How the Entrepreneurial Spirit Is Transforming the Public Sector by veteran city manager Ted Gaebler and journalist David Osborne. Gaebler and Osborne challenged one of the most common complaints about public administration -- that government agencies were irredeemably bureaucratic and resistant to change. The authors argued that that need not be the case. Government managers and employees could and should, the authors wrote, be as entrepreneurial as their private-sector counterparts. This meant embracing competition; measuring outcomes rather than inputs or processes; and insisting on accountability.

For public-sector leaders, Gaebler and Osborne’s book was a revelation. “I would say it has been the most influential book of the past 25 years,” says Robert J. O’Neill Jr., the executive director of the International City/County Management Association (ICMA). At the federal level, Reinventing Government inspired Vice President Al Gore’s National Performance Review. But it had its greatest impact on state and local governments. Public-sector officials across the country read Reinventing Government and ingested its ideas. Osborne joined the consulting firm Public Strategies Group and began hiring himself out as an adviser to governments.  

There’s no question states and localities function differently today than they did 25 years ago. Performance management systems, though not universally beloved, have become widespread. Departments and agencies routinely measure customer satisfaction. Advances in information technology have allowed governments to develop and share outcomes more easily than ever before. Some watchdog groups consider linking outcomes to budgets -- also known as performance-based budgeting -- to be a best practice. Government executives in many places talk about “innovation” as if they were Silicon Valley executives. This represents real, undeniable change.

Yet despite a generation of reinvention, government is less trusted than ever before. Performance management systems are sometimes seen not as an instrument of reform but as an obstacle to it. Performance-based budgeting has had successes, but they have rarely been sustained. Some of the most innovative efforts to improve government today are pursuing quite different approaches, emphasizing grassroots employee initiatives rather than strict managerial accountability. All of this raises a question: Has the reinventing government movement left a legacy of greater effectiveness, or have the systems it generated become roadblocks that today’s reformers must work around?  Or is the answer somehow “yes” to both of those questions?

Reinventing Government presented dozens of examples of “entrepreneurial” problem-solving, organized into 10 chapters. Each chapter illustrated a theme, such as results-oriented government or enterprising government. This structure -- concrete examples grouped around larger themes -- reflected the distinctive sensibilities of each author. Gaebler, as a city manager, had made a name for himself by treating constraints such as funding shortfalls or bureaucratic rules as opportunities. His was a bottom-up, let-a-hundred-flowers-bloom sensibility. He wanted his fellow managers to create cultures where risks could be taken and initiative could be rewarded.

Osborne, a journalist, was more of a systematizer, drawn to sweeping ideas. In his previous book, Laboratories of Democracy, he had profiled six governors who he believed were developing new approaches for delivering services that constituted a “third way” between big government liberalism and anti-government conservatism. Reinventing Government suggested how that would work in practice. It also offered readers a daring and novel vision of what government’s core mission should be. Government, the book argued, should focus less on operating programs and more on overseeing them. Instead of “rowing” (stressing administrative detail), senior public officials should do more “steering” (concentrating on overall strategy). They should contract out more, embrace competition and insist on accountability. This aspect of Osborne’s thinking became more pronounced as time went by.

“Today we are well beyond the experimental approach,” Osborne and Peter Hutchinson, a former Minnesota finance commissioner, wrote in their 2004 book, The Price of Government: Getting the Results We Need in an Age of Permanent Fiscal Crisis. A decade of experience had produced a proven set of strategies, the book continued. The foremost should be to turn the budget process “on its head, so that it starts with the results we demand and the price we are willing to pay rather than the programs we have and the costs they incur.” In other words, performance-based budgeting. Then, they continued, “we must cut government down to its most effective size and shape, through strategic reviews, consolidation and reorganization.”

Assessing the influence and efficacy of these ideas is difficult. According to the U.S. Census, the United States has 90,106 state and local governments. Tens of thousands of public employees read Reinventing Government and the books that followed. Surveys have shown that the use of performance measurement systems is widespread across state, county and municipal government. Yet only a handful of studies have sought to evaluate systematically the impact of Reinventing Government’s core ideas. Most have focused on just one, the idea highlighted in The Price of Government: budgeting for outcomes.

To evaluate the reinventing government movement primarily by assessing performance-based budgeting might seem a bit narrow. But paying close attention to the budgeting process is the key to understanding the impact of the entire enterprise. It reveals the difficulty of sustaining even successful innovations. The experience of Iowa illustrates the potential of a vigorous reinventing government initiative -- and the pitfalls of that approach.

It started in early 2001, when the U.S. economy fell into a recession. Iowa’s revenues dropped by 11 percent, the steepest decline in the country. The resulting budget gap was too large to close with incremental across-the-board cuts. It was clear to everyone that deep spending reductions and/or significant changes in the way government operated would have to be made. To help with this process, Iowa Gov. Tom Vilsack turned to the reinventing government playbook, and to the consulting firm for which Osborne worked, the Public Strategies Group (PSG).

The first priority was the budget. Instead of universal cuts, PSG suggested that the state try budgeting for outcomes. Rather than starting with the previous year’s budgets and then making incremental adjustments, PSG wanted state agencies to assume they were working with a clean slate and plug in their desired outcomes. This wasn’t a new idea. Government reformers had talked about “zero-based budgeting” since the early 1970s. However, few state governments had tried it. That didn’t deter Iowa Budget Director Cynthia Eisenhauer.

At first, agencies had a hard time figuring out what to do. When, for instance, the Department of Corrections submitted its budget request, it did what it had always done. It used the last year’s budget as its baseline. Eisenhauer asked the department head to do something different: Imagine, she told him, that you were creating your department from scratch. What would that budget request look like? The director said if given the chance, he’d spend more on substance abuse treatment and skills development. It might take a few years, but doing those things would ultimately allow the state to reduce its spending on incarceration, which was far more costly.

The corrections department got the funds to make this investment. Other state agencies, notably the departments of Transportation and Human Services, were led through a vigorous process of rightsizing. Field offices for health and child welfare were consolidated; middle management was pruned. Five large agencies were allowed to convert to “charter” status: They received greater autonomy to hire and operate. In exchange, they promised to deliver savings. It worked. The state had hoped to realize $15 million in two years from such reinventing government initiatives. Instead, it saved $22 million. By the end of year three, savings had reached $50 million.

But as time went on, it became clear that the reinvention effort had neglected to pay sufficient attention to an important actor -- the legislature.

“One thing that I believe we did not do well was engage legislators and the public in what we were doing,” says Eisenhauer. That led to some fights that might well have been avoidable. One was over executive branch efforts to improve the child welfare system. When the state sought to base payments to nonprofit agencies on outcomes (such as child safety) instead of processes (such as the number of children served), the nonprofit managers went to their legislators to protest. Legislators took their complaints to the governor’s office. The problem was that the legislature had not really been included in the reinventing government process.

“Reinventing government was relatively blind to the role of legislatures in general,” says University of Maryland public policy professor and Governing columnist Donald F. Kettl. “There was this sense that the real problem was that good people were trapped in a bad system and that freeing administrators to do what they knew how to do best would yield vast improvements. What was not part of the debate was the role that legislatures might have played in creating those constraints to begin with.”

Over time, a pattern emerged. During periods of crisis, chief executives were able to implement performance-based budgeting. Often, it worked. But eventually legislatures pushed back. Maine offers a striking case in point. In 2002, agencies began using a performance-budgeting system that was supposed to give them greater autonomy in exchange for providing greater accountability. But that’s not what happened, says Grant Pennoyer, director of the office that administers the state legislature.

Instead, he says, legislators were ultimately unwilling to give up traditional ways of thinking and focus on results. “They created programs, and they wanted to keep them and make sure they were funded.” So in 2008, the legislature rolled back the system and returned to incremental budgeting. These days performance measurement in Maine’s agencies seems like a hit- and-miss affair, says Beth Ashcroft, who heads Maine’s Office of Program Evaluation and Government Accountability. Ashcroft’s agency audits or investigates three or four agencies every year. “When we ask for performance goals and measurements,” she says, “there are very few agencies that have something ready to hand over to us.”

This points to another problem that has bedeviled reinventing government efforts -- sustainability. Again, Iowa is a good example. According to both the state and PSG, Iowa’s reinventing government initiative saved tens of millions of dollars. Yet when Chet Culver took over as the state’s new governor in 2007, he promptly rolled back reinventing government initiatives including the charter status previously granted to five state agencies. What was the rationale for rolling back such an effective reform? The new administration wanted the leeway to make across-the-board cuts.

Politicians and agency leaders like to talk about culture change. But in the public sector, locking in permanent change is hard. That’s by design, says Vanderbilt University political scientist David Lewis. After all, he notes, “The public sector has regular elections whose purpose at least in part is to determine what the purpose of government should be.” To further facilitate policy changes, it also provides for the entire upper management of government agencies to change regularly. “Take the highest executives, lop them off every four years and replace them with new people. That is not a recipe for sustainability,” says Lewis.

Even public-sector leaders who do enjoy greater job stability -- city managers -- have been reluctant to connect performance measurement with budgeting. In 2011, ICMA sent a questionnaire to city managers across the country. The questionnaire was written by Barry Feldman, the city manager in West Hartford, Conn. Feldman knew that many of his fellow city managers talked a big game when it came to reinventing government. Previous studies had shown that the use of performance measures was widespread among state, county and municipal governments. But were they connecting outcomes to budget requests? Were they budgeting for training and education? For investments that would pay off in years to come?

“What I was trying to find out,” says Feldman, “was whether we were talking about this stuff or actually doing it. Were city managers acting more entrepreneurially, taking risks like the private sector?” The answer, he found, was that most were not. City managers were paying lip service to performance-based budgeting, but not acting on it. “They talked,” Feldman says, “but they did not take the actions.”

Clearly, performance-based budgeting has been hard to sustain. What about Reinventing Government’s other ideas? Osborne himself offers a mixed review. “In terms of what got the most traction, clearly catalytic government [the idea that government should steer, not row], mission-oriented government, customer-driven government, decentralized government, all of those got a lot of traction,” Osborne says. Ideas that didn’t find a lot of support included enterprising government, the notion that government should sometimes operate profitable businesses, and community-owned government, the concept that government should enlist the people as partners.

That’s a take that Sandy Borins, a Harvard professor who charts organizational trends, largely agrees with. “There are a lot of aspects that did work and are still with us,” Borins says. “One is the focus on measuring and improving the quality of service.” Compare renewing a driver’s license today, he says, “to what the process once entailed. You can thank Reinventing Government for the difference.” Borins also identifies the emphasis on performance measurement as a productive legacy of what he and other academics call “the new public management.”

Yet talk to people who have run these systems and you find a surprising ambivalence. As the head of Maryland’s StateStat system, Beth Blauer was in charge of one of the country’s most acclaimed performance management systems. She believes these systems, when used properly, can offer executives valuable assistance in making policy decisions. At the state level, she sees it as “a way to de-silo government and get agencies thinking about enterprise-wide problems.” But she also believes that Osborne tried to do too much with performance-based budgeting. “Where I departed [from the Reinventing Government playbook] was around the budget,” she says. “The work we were doing was about working with the legislature, about dealing with interest groups and achieving regulatory reform. The budget played a big role in the process, but it wasn’t what was leading the way.”

There was another problem. Measuring results, insisting on accountability -- these were supposed to spur creative problem-solving. But in practice, says Blauer, “whenever the budget was invoked in performance conversations, it automatically chilled innovative thinking; it chilled engagement,” she says. Agencies got defensive. Rather than focusing on solving hard problems, they focused on justifying past performance. This was precisely what Gregoire discovered in Washington at the end of her second term.

For Gregoire’s performance management director, Korthuis-Smith, the degree of managers’ disillusionment came as a surprise. Many had come to believe “it was just about compliance,” she says. While it did provide for leadership and accountability, it did little to motivate the front-line employees who were actually working directly with customers and clients on a day-to-day basis. “People weren’t necessarily committed to it,” she says. “It wasn’t sustainable.”

A group of state agencies had begun to experiment with a different approach to these problems, a set of principles and tools adopted from the manufacturing world known as “Lean.” This was a move away from the familiar top-down concept of reinventing government, and toward its opposite. Lean started at Toyota as a process for identifying and eliminating waste. It encouraged assembly-line workers to identify problems -- even small problems -- and to suggest solutions. The role of the manager in this process was closer to that of a coach than to that of a performance evaluator. Korthuis-Smith thought this softer bottom-up approach might play well in state government. Gregoire embraced the idea, as did her successor Jay Inslee, who took office in 2013. Later that year, his administration rolled out a new framework: Results Washington. He turned to Korthuis-Smith to run it.

Results Washington has not abandoned performance measurement. There are specific goals around homelessness, recidivism, traffic fatalities, high school graduation rates and environmental cleanup. However, the flavor of Results Washington is very different from the system that came before it. Instead of making presentations to the governor the primary venue for accountability, Korthuis-Smith and her team have assembled five “goal councils,” each made up of 12 to 15 state agency directors. The councils meet monthly to review data, discuss strategies and collaborate on solutions. “Agencies and customers helped build and test this system with us,” says Korthuis-Smith. “It’s not just ours -- it’s theirs.”

David Osborne has now been in the business of advising governments for more than 20 years. Ted Gaebler is retired but still travels and speaks on reinventing government issues. Both take enormous pride in Reinventing Government. Both believe it changed the way the public sector works. Yet both recognize that it did not achieve their larger goal.

The fact that reinventing government never sparked a revolution puzzles Gaebler to this day. “Why didn’t more of my colleagues pick it up and run with it?” he asks. He thinks the answer may be that many public managers were simply too risk-averse.

“Ted and I wanted to change the world,” Osborne says. He believes the crusade ultimately was a victim of bad luck. In the beginning, he insists, reinventing government did restore trust in the public sector. “If you look at the survey data,” he says, “in the ’90s, trust in government was rebuilding from a very low point. And a lot of basic services got better. Then the politicians began to screw things up. Clinton and the Lewinsky affair, that was a big hit. Then Bush invaded Iraq and then we had the recession, and now we have the most polarized politics in my lifetime. Distrust is focused on the politicians. It’s not focused on service delivery. And it is even deeper than it used to be because the politicians are doing such a bad job.”

Despite their concerns about the current political climate, Gaebler, Osborne and their fellow reinventors retain a sense of optimism about the future. “All of the things related to data and technology are really astonishing,” says Hutchinson, who worked with Osborne at PSG and now heads Accenture’s state and local government practice. “It’s happening so fast. We are going to know so much more than we ever did and be able to reallocate resources in ways that we never have before. There are huge breakthroughs coming.”