The State of Federal Management

t a hearing on Capitol Hill in 1968, a member of Congress summarized then-Postmaster General Lawrence O'Brien's managerial circumstances as "no control over your workload . . . no control over wages and . . . limited control over workplace conditions." O'Brien was taken aback but agreed, adding that if his constraints had been put that succinctly earlier, "perhaps I wouldn't be sitting here."

O'Brien's remark is a reminder that management in the public and private sectors is alike in all the unimportant respects. Most obviously, managing in the public sector is conducted in the glare of politics. Michael Blumenthal, who ran Bendix Corp. before serving as Treasury secretary, said once that in the private sector if he took time off to play golf, he was complimented for effectively delegating responsibility. If he did the same in government, he was accused of shirking his responsibility.

The unspoken truth at the heart of all efforts to improve government management and efficiency is that government wasn't designed to be efficient. It was mostly designed to be accountable, to advance social goals and to avoid fraud, waste and abuse.

Yet interest in better performance has grown. If Americans seemed, until Sept. 11 at least, to answer the question, "What do you want the federal government to do?" with a simple "less," at least they wanted that less to be done better. The quest to devise measurable indicators of results was codified in the 1993 Government Performance and Results Act (GPRA), and it has been reflected in the Clinton administration's Reinventing Government initiative and the Bush administration's President's Management Agenda. It is having some effect. The state of federal management has improved and is improving, as data on outputs, if not social outcomes, gets better and managers are impelled to pay more attention to them.

That trend has been reflected in the Federal Performance Project, conducted by Government Executive over the past four years in partnership with the Pew Charitable Trusts and with, first, Syracuse University's Maxwell School of Citizenship and Public Affairs, and later, The George Washington University Department of Public Administration. Under the project, agencies were assigned letter grades in five management areas: finance, human resources, information technology, capital assets (when it was relevant), and a broader category called managing for results. The project tried to focus on outputs, an approach that won generally high marks from observers. At first, the five functions were given equal weight in grading the agencies, but in the third year the project switched to basing half the grade on managing for results.

Not all the interest in results that produced that shift comes with a generous spirit, for "bad management," like "waste," is a handy stick to be used by those who simply don't like particular government programs. Yet a wide range of interviews conducted for this article-with evaluators of programs and those evaluated, along with informed observers-uncovered a good will that would surprise many Americans. Managers are capable people, and they want to do better, which is less of a surprise than it would seem to most Americans. After all, if we may be suspicious of the public service yearnings of the big political contributors who want to be U.S. ambassadors abroad, it takes another kind of man or woman to want to run the Internal Revenue Service.

All of the top officials whose agencies have been evaluated by the Federal Performance Project were political appointees. But most were actually qualified for their jobs, in training or experience. James Lee Witt, who managed a dramatic turnaround at the Federal Emergency Management Agency, had run disaster relief for Arkansas before his 1993 appointment; none of his predecessors had any such experience. Kenneth Kizer, a doctor who had run California's Health Services Department, took over the Veterans Health Administration in 1994. Yet his sweeping changes hardly earned him a hero's medal; quite the contrary, he was denied renomination for a second four-year term in 1998. The Administration for Children and Families' Olivia Golden, the Federal Housing Administration's William Apgar and Doris Meissner of the former Immigration and Naturalization Service, were academic specialists in the material they came to manage. The IRS' Charles Rossotti had built a technology company.

Those committed leaders struggled to do well in the context of mixed and changing signals about their agencies' missions from their authorizers, especially Congress but also public constituencies. They managed in tight boxes akin to O'Brien's-with outmoded technology, fragmentary measures of performance, weak systems for connecting money to purposes, and rigid personnel systems with an age distribution that threatens droves of imminent retirements. Yet the attention to results was great and growing, and the constraints of those boxes were loosening.

EVALUATING MANAGEMENT

The rub in any evaluation of results is that it is easier to measure inputs than outputs, let alone outcomes in society, and the chain linking inputs to outputs seldom is clear. For instance, the nation wants fewer hungry people-an outcome-but the information the Agriculture Department knows best is how much it spends on food stamps-an input. It might be able to measure how many stamps it delivers-an output. Yet do more food stamps delivered indicate more people fed or more hungry ones who need food? Even some input measures can be misleading. Clean audits of an agency's financial systems, for instance, look like a sign of good fiscal management, but not if they are achieved only by applying brute force at audit time in manually working around deficient systems.

Moreover, while the evaluations that exist overlap, they are not directly comparable, so developing a complete picture of how well or poorly the government is doing is no mean feat. The most comprehensive evaluations are done by the Office of Management and Budget. Because OMB's reviews are connected to decisions about agency budgets, managers consider its efforts the most significant. That was not the case in the past, when OMB's evaluation efforts were regarded as box-checking exercises conducted by junior officials.

The first evaluation method used by OMB in the Bush administration is the scorecard for departments and large agencies produced under the President's Management Agenda. In scoring Cabinet departments, the effort differs from the Federal Performance Project, which focused on agencies, regarding most departments as sprawling confederations of very different organizations, with management processes too varied to assess as a whole. OMB awards red, yellow or green lights in five management categories: strategic management of human capital, competitive sourcing, improving financial management, expanding e-government, and integrating budget and performance. It gives each department two ratings, one for overall performance and one for progress in implementing each of the initiatives. It thus tends to focus on inputs and to compare performance to previous history, not outcomes.

The other OMB evaluation effort, known as Performance and Management Assessments, is more results-oriented. It addresses programs, not departments or agencies, and thus takes a smaller bite than the Federal Performance Project. Still, when scores can be compared, they are broadly similar, and, indeed, the OMB scorings occasionally refer to Federal Performance Project assessments. The Social Security Administration and the National Weather Service have scored highly, while the IRS and the Immigration and Naturalization Service (which has been split up and moved into the Homeland Security Department) have fared poorly. OMB begins with program purpose and design, then looks at strategic planning, program management and finally at results and accountability, with the last accounting for half the total score. In looking at results, OMB explicitly compares the program being scored with similar programs elsewhere; for the National Weather Service, that means asking how it compares to Navy and European weather forecasting organizations.

These evaluations tend to focus more on how agencies are being managed than on how employees on the front lines view their organizations' performance. But surveys and interviews suggest that the appetite for innovation is greatest at the top and the bottom of agencies-in leaders and in the newest professionals. In between, middle managers are more likely to be operating on autopilot, tending to view any change in terms of its impact on the budgets they defend, and to be beaten down by years of trying to manage in the tight boxes they inhabit.

A survey issued in 2000 by the General Accounting Office provides evidence for that view. In the survey, GAO asked more than 3,000 managers in 28 agencies how they perceived their agencies were doing at instilling a performance-based culture. Overall, the results were dreary. At 11 agencies, fewer than half of the managers perceived a strong top leadership commitment to achieving results. At 26 agencies, fewer than half perceived that employees received positive recognition for helping their agencies accomplish their strategic goals.

While more than half of the managers at 22 agencies reported that they were held accountable for the results of their programs, at only one-the Office of Personnel Management-did a majority report that they had the decision-making authority they needed to achieve their goals. On an agency-by-agency basis, the survey's results overlapped only in part with the Federal Performance Project or OMB's efforts. Where comparison was possible, the results were parallel. The Forest Service, IRS and the Health Care Financing Administration (now the Centers for Medicare and Medicaid Services) tended to be at the bottom of each set of rankings, with Social Security and NASA toward the top and the Federal Aviation Administration in the middle.

Not surprisingly, how officials whose efforts were graded in the Federal Performance Project felt about the process depended on their score, when the reports were issued and, especially, how exposed their agencies were politically. For Olivia Golden, the evaluation came at the end of her tenure and was a welcome chance to put her story together and tell it. For her and many other officials, any attention to government performance absent a crisis was welcome. Also, having worried that the assessment would be naturally biased toward "harder," more measurable services such as Coast Guard rescues or Social Security telephone inquiries, Golden was pleased to find a receptive hearing for the Administration for Children and Families' "softer" services.

For another social services manager, who asked not to be identified, the Federal Performance Project was both too much and too little. It was too much because he knew his agency was badly managed but could not persuade Congress to provide either authority or money to fix many of the shortcomings. Moreover, he thought, any review, no matter how well-intentioned, was bound to produce a "bad management" stick for use by those on Capitol Hill who simply did not favor government assistance to poor people. On the other hand, the report was too little because he felt he didn't learn much and instead spent considerable time getting the reviewers up to speed about his agency. He deemed similar reviews by the National Academy of Public Administration or GAO more helpful because they were deeper and more interactive, and resulted in specific recommendations.

For the INS' Meissner, all of the evaluations were less helpful than they might have been because they did not vividly capture a particular manager's context. In her case, not only was the agency's mission ambivalent and its congressional oversight complicated, but it was undertaking major reforms as it was growing as rapidly as the World War II Pentagon-hardly easy circumstances in which to manage.

For the Coast Guard's Capt. Mark Blace, the Federal Performance Project was a target of opportunity to begin to weave a strategic view into a very tactical agency culture. The Coast Guard volunteered to be part of a pilot test for the project, then was evaluated two years later. So agency leaders had an incentive to construct a vision and try to achieve it. That process helped the Coast Guard draw a map from inputs to outputs, rather than simply projecting forward from existing activities. For instance, officials realized they did a good job of tracking the distress calls the Coast Guard answers, but were less focused on the mission of avoiding rescues in the first place by promoting water safety.

It is, of course, impossible to parse out the effect of any evaluation process, but the cumulative effect of various performance measurement efforts, starting with the Government Performance and Results Act, has clearly made a difference. Maurice McTigue, a former New Zealand Cabinet minister who runs the Mercatus Center at George Mason University, which focuses on agency accountability, says his discussions with federal chief financial officers move quickly from what they disclose to how they plan and measure results. They are competitive, wanting to look good and do well, and so they are willing recipients of suggestions. As a result, McTigue and others say, the concept of accountability is shifting from how money is spent to what is accomplished in the public interest.

MANAGING...TO DO WHAT?

The art of managing for results in government involves bringing into harmony three sides of a triangle: goals, institutional capacity, and authorizing environment. For a manager, that means answering three questions: What am I supposed to do? What is my agency capable of doing? And, critically, what will Congress (and my superiors or constituencies) allow me to do, or at least tolerate my doing? For their part, private sector managers usually can take goals as given, and while they must deal with intrusions from the "authorizers" in government and occasionally among their boards or shareholders, most of the time they are left to try to build organizational capacity to meet goals. They can be preoccupied with one leg of the triangle.

In government, however, effective management is difficult unless the three legs of the triangle are in tolerable alignment. The critical element in that process is not personal leadership-though that matters-but whether or not an agency's goals and mission are clear and supported by its authorizers. The Social Security Administration got the only A grade in the first year of the Federal Performance Project, which was a tribute to its efficiency, especially in adapting new information systems to serve its clients. Yet its goal is clear and clearly supported by both its constituencies and Congress: Get information and benefits checks to people who are entitled to them. The Weather Service, graded tops in the 2001 survey, also has a clear mission, one it translated into "no surprises" and then into more specific goals for warning times about storms.

At the other extreme, the INS ranked at the bottom the first year, despite a doubling in its budget over the previous half decade. Its many problems began with a split in its authorized mission. On one hand, the agency was supposed to keep illegal immigrants out of the United States. But on the other, it was charged with helping legal immigrants to become citizens and receive federal benefits. So, too, the Forest Service is charged with both preserving public lands and producing resources from them.

Sometimes, the problem is that Congress and the public keep changing their minds about agencies' missions. The IRS, for example, seems to have a clear-cut job: collect taxes from those who owe them. Yet when the agency sought to close the gap between what was owed and what was paid, by employing performance measures to increase collections, its tactics earned it a beating on Capitol Hill in 1997. Congress prohibited the IRS from using those measures. But then, predictably, by 2002, members of Congress were looking at declining IRS tax collections and criticizing the agency for going soft on tax cheats, and by 2003 the incoming IRS commissioner made going after them his top priority.

The same fate befell the Veterans Health Administration when it sought to restructure itself to serve a client base that was moving south and west, and with ills more likely to be caused by poverty and old age than combat-and thus more likely to need clinic and nursing home care than long stays in hospitals. Yet closing VA hospitals removed local icons of government beneficence-icons that employed people in various congressional districts-so communities and their representatives fought the closures.

In other cases, the sheer complexity of the authorizing environment creates headaches for managers. EPA received a decent grade (B-) in the 1999 Federal Performance Project, despite its authorizing handicaps. Cobbled together out of dozens of pre-existing programs, it is overseen by more than 40 committees and subcommittees of Congress. That legacy has forced EPA to operate in rigid stovepipes-focused on air, water, toxins and other pollutants. Decentralization into 10 regional offices was both a blessing and a curse. It made for responsiveness in addressing local circumstances, but the regional administrators became barons in their own right-and ones who also had to act through the stovepipes, so coherent national initiatives were doubly difficult to implement.

PEOPLE, MONEY AND TECHNOLOGY

The second leg of the triangle, capacity, also is important in government. So the attention to management inputs-such as staffing levels-is hardly beside the point. As a veteran of the Clinton administration's National Performance Review put it: "When we recommended that agencies take on downsizing by cutting headquarters more than field operations, many couldn't do it. They didn't have the human resources tools." For public agencies, as for private corporations, the most important element of capacity is their people. But unlike their private sector counterparts, many public agency heads manage in the presence of strong unions, rigid personnel systems, and employees who are drawn to their jobs partly by a desire for job security.

For agency heads, managing the capacity leg of the triangle means being creative in working around rigidities in the personnel system. The FAA, where labor relations had been embattled since the air traffic controllers' strike of 1981, used personnel flexibility granted by Congress to negotiate a five-year agreement with its union in 1998. By 2002, three-fourths of FAA employees worked under a pay-for-performance system. The FAA's top managers, the equivalents of Senior Executive Service members elsewhere, forgo raises based on length of service and instead get cash bonuses of up to a quarter of their salaries if they meet performance goals.

More recently-and dramatically-the Homeland Security and Defense departments have been freed by Congress from many civil service constraints. Soon, more than half of the civilian employees of the federal government will not work under traditional civil service rules. Their managers now have much more latitude in rewarding performance. With authority, however, also comes accountability, and across the government managers already have more latitude in personnel practices than they actually use.

In addition to managing the human resources challenge, many effective agencies have become very creative in dealing with funding issues. The one bright spot by 2002 in INS's otherwise dreary management picture was on the services side, where offering faster processing of visas for a $1,000 fee enabled the agency to hire more people, do more training and acquire better technology. Likewise, the National Park Service's authorization to charge recreational fees enabled it to make a dent in its maintenance backlog. And after the Food and Drug Administration was authorized in 1992 to collect fees from pharmaceutical companies, it cut drug approval times in half by 1999.

Likewise, effective procurement, especially of information technology, is a vital part of building capacity. Using authority given it in 1996 to operate outside ordinary procurement rules, the FAA was able to buy 22 central computers for monitoring flights in just 18 months.

Most government IT managers have shown they can get the job done, but find it much harder to demonstrate the returns on IT investments. When the INS sent student visa approvals for two of the Sept. 11 hijackers several months after the attack, it provided graphic testimony to the agency's continuing information woes.

MEASURING IMPROVEMENT

As the Federal Performance Project went on, it attempted to assess whether agencies were showing success in achieving better performance. In 2002, the project revisited six agencies it had first graded in 1999. Three had improved, while the other three had worsened. The biggest success story was the FAA, though its improvement was tempered by its evident failures in the Sept. 11 disaster. It took, for instance, six years to draft regulations to require better training and higher standards for the private firms that were responsible for screening at airports. The IRS also was a success story. The blowup with Congress in 1997 gave its leaders the chance to reshape the agency. Rossotti replaced the agency's 33 geographic baronies with four business units aligned with sectors of the tax-paying populations.

The two worst failures were both mostly a case of "more of the same only more so." The INS, for instance, grew in budget from $1.5 billion in 1993 to $5.5 billion in 2002, and in personnel from 18,400 to 36,400. Yet there still were only 2,000 investigators available to track the estimated 9 million illegal immigrants already in the country. And the agency's enforcement mission remained 180 degrees removed from its service one. Even on the enforcement side, its real mandate was not what it seemed. "Prevent illegal immigration" turned out not to be license to pursue illegals. Rather, INS' real mandate was shaped by communities along the border, which wanted that border to be safe and controlled.

The Social Security Administration, tops on the Federal Performance Project's list in 1999, stayed atop the 2002 list, but its grade dropped from an A to a B. The reasons lie between the goals and authorizing environment legs of the triangle. SSA continued to do what Congress wanted and what it had the institutional capacity to do. It was faulted, though, for ducking the riskier task of engaging in the policy process. Like the FAA with regard to airport security, SSA was missing in action in the debate over how to ensure the financial future not just of Social Security but of the Americans who are served by SSA's programs.

WHAT AMERICANS WANT

Last summer, the second Volcker commission, formally the National Commission on the Public Service, outlined a bold agenda for measuring the past and imagining the future of the federal government:

• Organize the government into a smaller number of mission-oriented departments like Homeland Security.
• Get much better leadership in government.
• Enhance government performance.

Now, by contrast, the threat is more diffuse. So will be the ways of dealing with it. The war on terror will be fought with local police as much as federal armies. It will be fought multilaterally, with coalitions cutting across nations and across the public and private sectors, with the government providing carrots, sticks and, mostly, sermons. Private citizens will be much more engaged, and, as the intrusive security procedures at airports suggest, they may have to change their behavior much more than they did during the Cold War.

There were hints of new forms of public-private partnership in the Federal Performance Project assessments. The Food Safety and Inspection Service, for example, shifted from "poke and sniff" inspections to working with food producers to develop and monitor inspection processes. But again, progress was limited: The requirements of the new partnership were seen as threatening by inspectors, only a quarter of whom had college degrees.

The teething problems of the Homeland Security Department have been understandable in trying to assemble a large department from a number of pre-existing pieces with different organizational cultures. Yet it is not clear that in the end whether "homeland security" is much clearer a mission than "defense" or "health and human services." Nevertheless, the government may well be streamlined, as the Volcker commission recommended, into a smaller number of more mission-oriented agencies, of the sort that tended to perform best in the project's surveys. That is likely to happen less by conscious choice than by the force of competition, as agencies move toward their core missions, privatizing or otherwise shucking off ancillary activities.

The circumstances of the war on terrorism do offer opportunities to bring new talent into government, as the Volcker commission seeks, but doing so will require rethinking not just hiring but careers and the laws that shape them. A recent RAND study looked at the requirements of leadership across the government, for-profit and not-for-profit sectors. The study found not so much shortages of particular skills-though the government did report some such shortages, given its relatively low pay and constraints in reaching out for talent. Rather, all three sectors reported that the nation is producing too few future leaders who combine substantive depth with international experience and outlook. So, too, managers with a broad strategic vision in a rapidly changing world are in short supply.

The practices of existing organizations do not produce such leaders-quite the contrary. That is most striking in the federal government, where lateral entry from other sectors is almost nonexistent, except at the top. In the short run, the answer for government is to seek much more flexibility to acquire talent laterally and for short periods. The government will continue to be hard-pressed to compete for talented professionals, such as top-flight economists.

Over the longer term, dramatic changes in patterns of leadership in government will be required. Under such a system, talented people would pursue "portfolio careers": They would be government officials one year and private sector executives the next. But that vision runs smack into U.S. procedures for political clearance in the short run and conflict-of-interest legislation in the long term. The logic of the future suggests that those conflict-of-interest provisions are outmoded, but that may not be the logic of politics-at least not soon.

The third Volcker commission recommendation was to develop new ways of enhancing the performance of the government. If agencies were more tightly organized around missions, their success would be easier to measure. The future seems likely to be "government by market," but of a special kind, with the emphasis on the "best provider." As the Volcker commission emphasizes, the federal workforce has shifted from clerks to specialists. Functions will move from public to private, to partnerships between the two, and back again.

It will be more and more necessary to spell out why government does much at all. The Social Security Administration earns high marks for customer service and getting checks out, but lots of private companies also do those tasks well. SSA is efficient at serving customers, but then so is Wal-Mart, and Wal-Mart comes with potential "offices" almost everywhere. If Disney runs theme parks, why not national parks, where so much is outsourced already?

The federal civil service of the future will be smaller in people, if not budget, with much more flexibility in hiring and rewarding those people. The lines separating the civil service from private employment will blur, as not just people but functions move from one sector to another. But procedural and legal constraints will remain. How formidable they are will depend, ultimately, on how distrustful Americans are of all large institutions, public and private.