By Frank Ostroff
When Hurricane Katrina engulfed New Orleans in the summer of 2005, the deaths, injuries, and damage to property that resulted were stark reminders of the cost to all of us when government at any level—federal, state, or local—does not perform as well as it should. The year before, the 9/11 Commission found that government’s failures to anticipate and respond to the terrorist attacks on that date were “symptoms of the government’s broader inability to adapt how it manages problems to the new challenges of the twenty-first century.” Although many public servants performed heroically, these horrific events and their aftermaths dramatize the need for high performance from government agencies both in dealing with life-and-death situations and in preventing crises from ever reaching that point.
This is a truth easily overlooked when the private sector is making impressive gains in productivity and discovering market solutions to large social needs. In reality, high-performing government agencies do resemble well-run companies. Both have worthy goals; well-designed, rational processes; strict accountability; and effective leaders. But the profound differences in their purposes, their cultures, and the contexts within which they operate conjure up quite different obstacles. The greatest challenge in bringing about successful change and significant, sustained performance improvement in the public sphere is not so much identifying solutions, which are mostly straightforward, as working around four unique obstacles.
First, agency leaders are not ordinarily chosen because of their commitment to spearheading reform or because they have a track record in leading large-scale change efforts. Rather, they are appointed on the basis of their command of policy, technical expertise in the agency’s work, or political connections.
Second, once a person is selected to lead an agency, he or she usually has only a limited amount of time to see a change effort through. The nomination process can occupy the first nine months or more of a U.S. president’s four-year term, and by the last year of that term, the agency head may already have begun looking for his or her next job. As a result, the average tenure of political appointees is effectively 18 to 24 months, tempting top agency officials to concentrate on policy reforms that can be enacted quickly, instead of on time-consuming organizational revampings whose results they may not still be around to see.
Third, rules governing such areas as procurement, personnel, and budgeting, which were originally adopted to prevent public-sector wrongdoing, have created workplaces that are significantly less flexible than those in the private sector. And legal doctrines intended to keep agencies’ activities within the scope of the powers delegated to them by Congress can inhibit initiative. Public-sector managers know, too, that the penalties for failure are almost always greater than the rewards for exceptional performance.
Finally, in a democracy, everyone has a rightful stake in an agency’s activities. Important constituencies include not only the president of the United States, cabinet members, members of Congress, and oversight organizations such as the Office of Management and Budget, but public-interest watchdog groups and the media. Most of an agency’s operations are conducted in a fishbowl, and almost every initiative is bound to meet with someone’s disapproval.
These facts of public life may never go away. But there are agency leaders who have figured out how to court support among key stakeholders, rededicate employees to an agency’s true mission, undertake reform so comprehensively that resistant elements are unable to subvert it, and lay the groundwork for next steps so clearly and systematically that progress continues when leadership changes hands.
The transformation of three federal organizations discussed below demonstrates how deep change and significant performance improvement can be achieved at public agencies. The Occupational Safety and Health Administration, or OSHA, which oversees workplace conditions mostly in the private sector, redefined its mission and goals and envisioned a new way to achieve them. The Government Accountability Office, or GAO, which investigates other federal agencies and issues reports on their performance, adopted many of the talent-management practices found in the private sector. And Special Operations transformed itself from an ad hoc arm of the military into an elite standing force comprising servicemen and servicewomen drawn from several military branches. It also boasts the military’s first command responsible for all Special Operations Forces.
Virtually every administration in the past 40 years has launched initiatives to improve government performance, including those of President Bush and President Clinton. On the basis of my experience as a consultant to both public and private sector organizations, I have identified five principles that characterize successful public-sector change efforts and can achieve the desired results.
Principle 1: Improve Performance Against Agency Mission
Public-sector organizations aren’t created to maximize shareholder wealth. Rather, they are charged with promoting a particular aspect of the public’s welfare. Effective and efficient execution of their mission is what taxpayers pay for. It’s also what motivates agency staffers. The reason most OSHA employees get up and go to work in the morning is to protect the safety and health of American workers. But mission can get blurred or lost as political priorities shift and agency leaders come and go. Even in the best of situations, mission is subject to varying interpretations.
When Joseph Dear became the assistant secretary of labor (and head of OSHA) in 1993, OSHA measured success chiefly in terms of the number of inspections conducted and fines imposed. While in certain situations inspections and fines were the appropriate response, they were not the only, and sometimes not the most effective, way of advancing OSHA’s mission. Clearly, the agency had become a captive of metrics originally intended to promote workplace safety but that had over time become an end in themselves.
How does drift like that occur? And why don’t leaders correct course when it does? In OSHA’s case, staffers’ exposure over the years to workplace injuries and fatalities that could have been avoided had instilled in some of them a punitive attitude toward business. The agency’s emphasis on inspections and fines had reinforced that attitude, preventing many employees from realizing that better alternatives might exist. Understandably, OSHA’s disciplinary approach antagonized many employers, who often underestimated the cost of workplace hazards to their employees and themselves. These businesses adopted the attitude, “the less OSHA does, the better.” To both groups, enforcement appeared to be a zero-sum game.
Many agency employees, however, don’t pick a side. They instead feel estranged from their agency’s strategy and mission. They don’t see how their individual efforts directly affect the agency’s performance, and so they start to focus on producing outputs, which are easy to quantify, rather than on achieving outcomes, to which private-sector measures like return on invested capital, or ROIC, do not apply. As employees lose sight of the overall mission, they may eventually come to care only about those things they can directly control, like protecting their own turf.
Accordingly, OSHA’s transformation effort began with a rededicated commitment to mission. That meant helping employees rediscover the reason the agency was created—to reduce the number of injuries, illnesses, and deaths in the workplace—and then reaching beyond it by calling for the elimination of all preventable workplace ills in ten years. Although literally impossible to achieve, this stretch goal was intended to stimulate innovative thinking. It also had the effect of making the agency’s reorientation impossible to doubt.
Once a mission has been articulated, agency leaders must put a stake in the ground by establishing improved performance against mission as the fundamental objective of the transformation effort. Doing so entails choosing clear performance-improvement goals and formulating specific initiatives. In the process, performance or skills gaps in the organization will be exposed. When David Walker became the U.S. comptroller general (and GAO’s leader) in 1998, the office’s ability to perform its mission had been damaged by a downsizing. Shortly after assuming his new position, Walker made addressing personnel and skill gaps a priority. In the case of Special Operations, aligning performance and mission meant adapting to the new reality of “asymmetrical warfare,” in which the enemy was symbolized not by a Russian tank and its crew but by terrorists on a commercial jet loaded with passengers and fuel.
In the business world, considerations like ROIC help companies set priorities and evaluate initiatives. Improving performance against mission is a framework for doing the same thing in public-sector organizations. OSHA used performance improvement goals to determine which initiatives should be undertaken. For example, OSHA’s Atlanta East office obtained a commitment from a local steel company to provide all its workers with equipment designed to protect them from falls. In the first six months that the agreement was in force, three workers fell from heights of 60 feet or more. Without the equipment they were wearing, all three would have died. Over the same period, workers’ compensation claims at the company went from more than $1 million to $13,200; in the first three months, accident costs per person-hour dropped by 96%.
Principle 2: Win Over Stakeholders
Whereas CEOs have to please such constituencies as lenders, securities analysts, and shareholders, the range of stakeholders that agency heads must cultivate is even wider. Broadly speaking, they fall into two groups—external and internal.
Special Operations Forces were active during the Vietnam War, operating behind enemy lines and in combination with indigenous forces, but had been nearly put out of business after the war’s end. The army reduced Special Forces from seven active groups to three, the navy cut the number of SEALs by half, and the air force deactivated all its Special Operations gunships.
As a result, the United States lost most of its ability to launch and sustain demanding, clandestine operations in support of conventional U.S. forces. The loss was most apparent in the failed 1980 attempt to rescue American hostages in Iran. A group of soldiers, Defense Department officials, and members of Congress and their staffs became very concerned about the United States’ lack of preparedness in the face of terrorism, foreign insurgencies, and other unconventional threats to national security. This group launched a campaign to revamp Special Operations to address these dangers, leading to the passage, in 1987, of the Nunn-Cohen Amendment to the Goldwater-Nichols Department of Defense Reorganization Act of 1986, which created the first command responsible for all Special Operations Forces. Headed by a four-star general, the reconstituted SOF now includes Army Rangers, Air Force Special Operations, and Navy SEALs as well as Marine Special Operations Command units.
Even after the joint command was established, the SOF leadership had to convince current stakeholders of the range and value of the forces’ capabilities. One way it did this was by inviting senior military officials and political leaders to Fort Bragg to observe the soldiers as they went through their exercises. According to General Wayne Downing, SOF’s third commander, soldiers were encouraged to relate the breadth of their experience and expertise, including their mastery of foreign languages and cultures. On one occasion, he recalls, a soldier spoke about the medical care and education she provided to tribes in the hills of Oman.
The SOF leadership also wanted U.S. diplomats to understand how Special Forces could be helpful to them. During a six-week training course, new ambassadors were invited to fly to Fort Bragg. On the flight was a platoon of SEALs, dressed in combat gear. The SEALs held a briefing and then put on parachutes. As General Downing described it: “We drop the tailgate of the airplane, and then the SEALs go out of the end of the plane. When the ambassadors get off the plane, the SEALs are waiting for them. That grabs their attention. We show them psychological operations, shooting, hostage rescue situations.” Several of the ambassadors, for example, took the role of hostage in a training exercise. An assault team gave participants a demonstration at the firing range. The SEALs also conducted a night mission. “After seeing us,” Downing continues, “[the ambassadors] can understand what we do and how we might be helpful to them.”
GAO takes its own approach to winning over stakeholders. “Theoretically, I have 535 bosses in the Senate and the House,” Walker says. “I respect them all but have to concentrate on the ones with the most interest in an issue. We identify stakeholders and work hard to understand their issues and concerns. If the issue is job classification, for example, I focus on the chairman and ranking member of the congressional committee with jurisdiction over that issue, as well as on members with local [Washington, DC] interests, since 75% of GAO employees are based in DC.” His attention extends beyond supporters. “Since we are a public agency,” Walker explains, “the potential opposition knows what we are trying to do early on. This is why it is so important to get ahead of the curve—to know the issues and then meet them ahead of time.”
Public-sector employees often stay at their agencies for a long time, typically much longer than their agencies’ leaders. And many have watched change efforts come and go—to little effect. But staffers’ longevity can actually be helpful to a leader seeking change. That is because those employees know a lot about how their agencies run and where they falter. By actively eliciting operational knowledge from them, leaders not only lay the intellectual foundation for the change effort, they also help gain the employee support needed for it to succeed.
In my experience, at any given agency, about a quarter of employees are initially receptive to a change initiative (sometimes out of frustration with how things have been handled in the past), a quarter are resistant, and the remaining half are on the fence. The continuing receptivity of the first group cannot be taken for granted. To keep those employees on board, the goals of the change effort must be consonant with their values—the reasons they came to the agency in the first place. The articulation of a stretch goal like OSHA’s—“eliminate all preventable workplace ills in ten years”—helps demonstrate the sincerity of the new leadership’s commitment, even in the eyes of the doubters. Goals like “centralize IT” or “reduce management layers,” by themselves, will not generate the amount of energy necessary to transform an agency’s way of working and view of itself.
Goals like “centralize IT” or “reduce management layers,” by themselves, will not generate the amount of energy necessary to transform an agency.
Questionnaires, interviews, and observation can determine who in the organization is amenable to change. Lack of change readiness can usually be attributed to issues of skill and will. Some may doubt their ability to keep up in the new organization. They think, “I’ve been successful at my job for 20 years, but I’m afraid I don’t have the skills to succeed in the organization being proposed.” Others may lack the will to engage in yet another change effort: “I don’t believe the proposed changes will improve performance.” Or, “The proposed changes threaten my turf.” Or even, “I just don’t have the motivation to cope with so much change.”
Well-crafted training programs can allay concerns about skill deficiencies. Their value is both psychological and practical. As employees gain confidence, they become more open to changes in their work or environment. Other tactics address a lack of will. OSHA, for example, convened a “diagonal slice” change team representing all agency functions and reporting levels, as well as both management and union members, to develop employee understanding of the agency’s performance challenges and support for recommended changes. The team visited high-performing public agencies and companies to learn from their experience in combating workplace ills. Accompanying the team were some people who had initially opposed the change effort, but were chosen in the hope that what they saw on the visits would help soften their resistance—and it did. The State of Georgia’s Environmental Protection Division, the team learned, allocates its limited resources by pinpointing the state’s environmental hot spots. And the Argonaut Insurance Company, it found, quantified the cost to businesses of workplace injuries and then helped those businesses implement safety and accountability systems. From this sort of exposure, the team members gained a sophisticated grasp of best practices and, no less important, a newfound belief in their feasibility.
To encourage GAO staffers to embrace new procedures, Walker focused on incentives. GAO had been a place where almost all employees received pay increases largely on the basis of time on the job and job classification or grade, regardless of performance. Now, compensation is structured on market-based salary ranges, and employees are rewarded for expertise, leadership, increased responsibility, and other contributions to performance.
Principle 3: Create a Road Map
In the mid-1980s, MBA graduates seemed to regard manufacturing as a black box: You put some things in and out pops a product at the other end. Many government reformers view the transformation process in similar fashion and hence fail to pay careful attention to the steps necessary to get from “here” (current agency status) to “there” (improved performance).
A change effort road map generally has three major phases: identify performance objectives; set priorities; and roll out the program.
Identify performance objectives.
In any change effort, you need to start at the top and then quickly move to ensure participation and support of a broad cross-section of employees. It is the prerogative of the agency leader and his or her senior managers to define the mission. At GAO, for example, David Walker began by talking with Congress and the agency’s two key internal groups—the agency’s managing directors and the 25 employee representatives who sit on the Employees’ Advisory Council. “We talk about what we need to do. I discuss it with them live so that they can provide input and ask questions.”
The agency leader can then commission a change team composed of individuals who are highly respected by agency peers, strongly support the need for change, and represent the various areas of the agency directly affected by the change effort. This team identifies the areas of performance requiring the most urgent attention and outlines the biggest obstacles to reform.
One way for the change team to do these things is to conduct internal fact-finding through interviews with senior managers, headquarters staff, field personnel, and outside experts. The team might also review internal reports, congressional oversight committees’ documents, and articles and books by experts who have studied the agency. The team should analyze past change efforts to determine which had positive effects, which were shrugged off, and why.
The change team can then hold redesign workshops to develop recommendations for improving performance. In OSHA’s case, strategy, organization, and process redesign workshops were conducted to develop a model for a new, higher-performing field enforcement office. One such workshop concerned the handling of informal complaints—that is, those reported orally. A map of the current complaint process was placed in the front of the room where the workshop was held. A facilitator briefly outlined the current procedure and described steps, based on best practices, that could be taken. The facilitator then asked the group a series of questions such as, “Does the process have any redundant steps?” “Are there handoffs that should be eliminated?” “Are there steps that should be added?” “Which ones should be automated?” For every step of the current process, the workshop participants, which included members of the change team and OSHA employees who had either handled informal complaints or were familiar with the process, came up with suggestions for improvement. In the course of one afternoon, some 150 ideas were generated.
Once all the suggestions are on the table, the next step is to decide which to adopt and in what sequence. Should an agency concentrate on areas where the potential for improvement is most marked? On areas that external stakeholders, including the general public, care most about? Or on areas where one can get results the fastest, thereby inspiring further efforts?
For most programs, I recommend constructing a 2 x 2 matrix, indicating high and low impact on performance on one axis and high and low difficulty of implementation on the other. One would almost always recommend immediately implementing those ideas likely to have the greatest impact on improving performance against mission while posing the least amount of difficulty.
Of course, there are times when it’s clear that an initiative will have such significant impact on performance that the hardship involved in getting there should be discounted. In using a similar matrix to decide which of the workshop’s ideas to implement first, OSHA found that implementing certain process redesigns was going to entail five weeks of staff training, which represented time away from inspection and enforcement. But the programs proceeded because both the change-team and agency leadership believed they would have such a significant effect on performance that the cost was justified.
It also helps to pursue tangible results that can be achieved quickly, even if they are not the ones that will have the biggest impact. At every one of OSHA’s approximately 65 field offices, frontline compliance officers were given a menu of improvement opportunities and then asked to pick the one they thought was most urgently needed. They then had to commit themselves to achieving extremely precise performance goals within eight weeks. This technique, called “breakthrough,” had startlingly good results. At OSHA’s Parsippany, New Jersey, office, for example, response time to employee complaints of serious hazards was cut in half after only eight weeks. OSHA’s breakthrough initiative demonstrates that meaningful change does not have to happen slowly. Quick wins also help generate faith in change efforts that unavoidably take years.
Roll out the change program.
It’s critical that agencies sow the seeds of change in fertile ground. Because Parsippany and Atlanta East were the OSHA offices judged the most receptive to change, they became the first pilot offices. Staff members of those offices were made virtual members of the change team (which was based in Washington), helping to ensure that the ideas the workshops recommended were suited to implementation in the field. An orientation and training plan was then developed, and risk controls were put in place. Representatives of the change team were present for the first month or so to provide guidance, solve problems as they arose, and discover what worked well and what didn’t.
Upon completion of the pilot phase, implementation was extended to five more offices, which were given three months to adopt the changes deemed most likely to improve performance against mission. To help keep the rollout on track as it spread to more offices, each successive office would have on the premises a couple of observers from one of the five offices next on the list. These observers would then help lead implementation at their own offices, along with members of the change team and veterans of the previous round. Rollout to all of OSHA’s field offices took three years, during which there was a change in agency leadership. Because the rollout had gained broad-based employee support, gathered momentum, and was already showing results, OSHA was able to surmount this usually disruptive event and achieve its goals.
Managing the design and rollout of a change program requires the involvement of a steering committee usually composed of the agency’s leader and senior managers of areas particularly affected by the transformation. The committee approves the sequence of steps, imposes milestones (for both process stages and real-world outcomes), specifies deliverables, approves change-team recommendations, and defines the expected contributions from both work units and individuals. The committee also takes ultimate responsibility for guiding the initiative and intervening to correct course when necessary.
Principle 4: Take a Comprehensive Approach
By now it should be clear that there is more to organizational redesign than moving boxes around a chart. For organizations to perform at a superior level, the full range of factors—leadership, structure, processes, infrastructure (including technology), people, and performance management—must be integrated and aligned. Yet the tendency within government is to seize on whichever organizational element the particular person or group driving the change effort knows best, at the expense of other elements.
The intense demands placed on Special Operations required a holistic approach to change. The transformation of Special Operations encompassed all the broad areas that must be addressed. Introducing a new unified command made up of generals from each military branch and headed by a general reporting directly to the secretary of defense was among the changes in how special operations would be led. Assigning special operations elements from multiple military services to the new organization and having them report to the unified command were among the structural changes. To obtain needed equipment, SOF created a much faster, flexible, and cost-efficient procurement process. Improved technology and weapons systems such as the laser designators used to pinpoint Taliban targets in Afghanistan and remote-controlled Predator UAVs (unmanned aerial vehicles), which conducted valuable surveillance of the Taliban’s movements, were adopted.
According to General Richard Potter, the efficacy of these upgraded elements depends entirely on the caliber of the troops themselves. SOF has therefore placed unprecedented emphasis on recruitment standards and training. The general explains, “For Army Special Forces, we carefully prescreen candidates and look for the attributes critical to succeeding as an SOF warrior. One of the goals of the training [that follows] is to strip off the veneer and see the inner man. We put the soldiers through sleep deprivation, intense psychological and physical stress, and demanding intellectual problems. After that, we send the soldiers off for individual skill training—weapons, medical, operations, intelligence, field operations, language and cultural skills, and negotiation. This whole process can [take] 1.5 to two years.”
Adopting a comprehensive approach may even require integrating activities across organizational boundaries. General Downing explains why it was necessary in SOF’s case: “Let’s say there is a camp containing terrorists that have killed Americans that we want to target. The Navy SEALs will provide reconnaissance of who is in the camp and when. The Army Rangers will attack the camp and kill or capture the terrorists. Air Force Special Operations, operating in tight coordination with the mission, can then fly in special planes and extract the terrorists and our Rangers. This requires very tight coordination and integration between these units.” The need for integration and improved performance was a lesson taken to heart, and acted upon, after the failed Iran hostage rescue mission. “In the opening days of Operation Enduring Freedom in Afghanistan,” recounts General Doug Brown, current SOF Commander, “U.S. Special Operations Forces successfully conducted 23 missions that were longer in duration, over greater distances, and more complex than Operation Eagle Claw [the attempt to rescue American hostages in Iran].”
In some situations, it may be difficult to overhaul all elements affecting performance at once. But even in the course of tackling the most pressing ones, it’s important not to neglect addressing the other elements altogether.
Principle 5: Be a Leader, Not a Bureaucrat
We’ve now established what it takes to lead a change program. Formulate a vision. Be aware of present realities. Develop a broad base of support. Set a clear path. Respect the complexity of what you’re attempting. Hold people accountable for both results and commitment to the change effort.
There are, however, two qualities of public-sector leaders that make such work difficult. First, it is in the nature of bureaucrats to respect barriers. Change leaders don’t necessarily knock them over; instead they find ways to see over and around them. As Walker puts it, “I find that often you have more flexibility than people believe. Many rules, as well as civil service limitations on what you can and can’t do, are good, and they need to be followed. But there is a difference between what you can and can’t do and what has been done and not done in the past.” As reported by GAO, during Walker’s tenure, that agency has roughly doubled savings achieved and resources freed up from $19 billion per year to $40 billion at other agencies as a result of its recommendations.
General Downing provides an illustration of how Special Operations has worked around barriers to obtain the equipment SOF needs. “Bringing complicated equipment online often takes ten to 15 years. We needed a new speed boat. Rather than going through traditional military procurement procedures, we used an innovative approach, having industry vendors build three different prototypes. After a thorough competition, we selected the best one. We had the first Mark V in 37 months.”
The other problem many agency leaders face is the perception that because they are political appointees, their commitment to improving performance against mission may be questionable. Such leaders must convince stakeholders of their sincerity. Agency employees mostly start out believing in the agency’s mission, which, whatever its particular focus, involves serving citizens and taxpayers. Over time, they see change programs come and go without making a dent. Meanwhile, the public interest is neglected. If an agency head can convince the rank and file that this time is different—that he is committed, is willing to invest the personal time and energy that is required, and will commit the necessary people and resources—then its original dedication will be reawakened.• • •
Many corporations have a deeply felt sense of mission over and above pleasing customers and enriching shareholders. Employees of pharmaceutical companies, for instance, are motivated to help cure illnesses. At most government agencies, such larger purposes are their entire purpose. When these objectives are misconceived or unclear, however, the agency’s activities lose their point. The dramatic changes at OSHA, the Government Accountability Office, and Special Operations Forces clearly show that change at public agencies is possible. Attesting to SOF’s successful transformation, Defense Secretary Donald Rumsfeld told the Wall Street Journal in February, “Instead of…operating generally only in support of someone else, we would have situations where Special Operations Command might be the one supported by other commanders around the globe.”
Public agencies can be mysterious places. But the solutions to reforming them are not. What’s required is a recognition that successful change is possible and that a proven set of techniques is available to get you there. Agencies with the vision and courage to undertake meaningful change can use these five principles to achieve their highest purpose.
A version of this article appeared in the May 2006
issue of Harvard Business Review