Command Performance

ritain's Driver and Vehicle Licensing Agency (DVLA) used to be so badly run that it was a national joke. The agency, which handles millions of vehicle registrations and driver's licenses every year, sometimes took months to process applications. Drivers boasted they could go for years without DVLA noticing their renewal fees hadn't been paid.

That changed in 1989, when DVLA became Britain's first "executive agency." It was given more freedom to manage money and people. In return, the agency was expected to meet tough targets for service quality and compliance with the law. DVLA's new chief executive was paid a bonus based on meeting the targets.

DVLA's transformation has been remarkable. The agency adopted new technologies, reengineered work processes and rationalized field offices. Turnaround time for applications dropped dramatically, and the agency won an award for excellence in customer service. Compliance among drivers also improved. Moreover, DVLA achieved this with fewer resources: Its budget dropped by 10 percent between 1991 and 1995 and staffing fell by 15 percent.

Success stories like this inspired Vice President Al Gore's National Performance Review to propose similar reform in the U.S. government. In March, Gore unveiled a plan to convert service delivery activities to "performance-based organizations" (PBOs), that will be given more freedom in exchange for tougher accountability for results. By this fall, bills to establish several PBOs will be sent to Congress. More may follow later.

David Osborne, co-author of Reinventing Government, says the PBO plan holds enormous promise. The British experience seems to support that conclusion. But the British have also encountered big challenges in implementation that will probably be repeated here. And Congress has concerns about the impact of the PBO plan on its traditional oversight role.

Britain's 'Next Steps'

The British reforms that are the model for the PBO plan came out of a 1988 study by Sir Robin Ibbs, a senior adviser to Prime Minister Margaret Thatcher. Most work within the British civil service, Ibbs said, consisted of delivering services to citizens. But that didn't mean the civil service was good at it. Many managers lacked the skills needed to improve service, and there was little pressure to improve results. Furthermore, Ibbs argued, there were too many central rules that made it difficult for managers to improve service quality.

Ibbs' proposal, known as the Next Steps plan, was to dramatically change structures and incentives within the British civil service. Parts of government that were concerned mainly with service delivery would be set up as executive agencies that would be given more flexibility to do their work. Each executive agency would negotiate with its parent department an annual performance agreement include measurable targets for financial performance, efficiency and service quality. At the end of the year, each agency would report whether it had achieved those targets.

Ibbs also proposed new rules for hiring and paying the heads of executive agencies. Chief executives would be recruited through open competitions and appointed for a limited term, usually three years. They were to get yearly bonuses of up to 20 percent of base salary tied to their success in meeting annual performance targets.

The first executive agency, DVLA, was set up in 1988. By April 1996, there were 125 agencies, with 37 more candidates being considered. Officials expect that by the end of 1997, 75 percent of the British civil service will work in executive agencies. Agencies vary substantially in size, from the massive Social Security Benefits Agency, with a staff of 67,000, to the Wilton Park Conference Center, with a staff of 30. One-quarter of the executives recruited since 1988 have come from the private sector. Some of the old agency chiefs won the new chief executive positions.

Many observers credit the Next Steps plan for achieving large improvements in efficiency and service quality throughout the British civil service. The Social Security Benefits Agency, for example, now handles a workload that would have required 15,000 more employees in 1990.

The author of one study suggests that the publication of performance targets increases pressure on managers and sharpens legislative oversight of the agencies. "There is a clarity of accountability that is very real," DVLA's first executive, Stephen Curtis, told legislators in 1994. "It does make a difference." Curtis also credited the "genuine managerial freedom" that has been given to agencies.

The Next Steps reforms received support from some unexpected places. John Smith, former leader of the opposition Labor Party, endorsed the initiative in 1991. However, Labor's support has weakened as the Conservative government puts more emphasis on privatization and contracting out. A parliamentary committee reviewed the reforms in 1994 and called them "the single most successful civil service reform program of recent decades."

How The U.S. Plan Developed

The decision to pursue in the U.S. reforms based on the UK model was made last summer, after a meeting between Vice President Gore and Administration officials and David Hunt, the minister responsible for civil service in the British government.

The PBO plan marks a change of strategy for the National Performance Review (NPR). In its 1993 report, NPR proposed several reforms in laws governing the civil service, procurement, budgeting and administrative services. Although NPR has achieved important progress in some areas like procurement, many other governmentwide proposals haven't been adopted by Congress. The PBO plan, by contrast, takes an incremental approach. Flexibilities will be proposed for individual organizations, rather than government as a whole.

The decision to move ahead with PBOs was motivated by last year's debate over the future of the Commerce Department. The first PBO candidates come from Commerce. The PBO plan can be regarded as one element of the Administration's response to complaints that Commerce ought to be dismantled or largely privatized. The plan would give units within the department more independence in operational matters but would retain a policy-making and oversight role for Commerce's core staff.

Vice President Gore announced last September that Commerce's Patent and Trademark Office would be the first PBO candidate. In March, Gore revealed the names of seven more candidates, including the $6 billion-a-year Defense Commissary Agency. Additional candidates are also being considered. For example, NPR's 1995 report included a suggestion that the National Oceanic and Atmospheric Administration's Mapping and Charting Service be made into a PBO-style organization. And Occupational Safety and Health Administration chief Joseph Dear recently said that his organization might volunteer for performance-based organization status.

The PBO candidates announced so far account for only a small portion of the government's 2 million employees, but advocates say there's lots of room to expand the plan. John Koskinen, OMB's deputy director for management says "there are significant areas of the federal government that provide services which are stable and measurable and would benefit from the application of the PBO concept." John Kamensky, NPR's deputy project director, points out that roughly two-thirds of federal employees deliver services directly to the public or customers within government. Direct service agencies include large organizations such as the Veterans Affairs Department and the Social Security Administration.

How Much Freedom?

One of the two key proposals of the PBO plan is more freedom to tailor internal activities to suit the goals and circumstances of the organization. Commerce's National Technical Information Service (NTIS), a PBO candidate which provides online services for a range of government agencies, wants to speed up procurement for rapidly evolving computer systems. NTIS' director, Don Johnson, argues that personnel and budget flexibilities will allow the organization to serve clients more effectively.

One of the first challenges in developing the PBO plan has been deciding how much freedom to give PBOs, and what form that freedom should take. Managers in some PBO candidates have pushed for broad, legislative exemptions that would give flexibility in a range of areas, including personnel management, procurement, real property management, budgeting, fee-setting, staffing limits, printing and paperwork requirements. The aim is partly to escape laws that tie up line operations. However, some managers see such exemptions as a way of circumventing central agencies, such as OMB and OPM, that are perceived as slow or stingy in giving waivers under existing law.

For some managers, the model for reform was provided in the 1996 Transportation appropriation. The law, signed last November, gave the Federal Aviation Administration the freedom to design its own personnel and acquisition systems. A similar approach was taken in HR 2533, the administration's bill to establish the Patent and Trademark Office as a PBO. The bill was sent to Congress last fall.

This approach can generate strong opposition. The FAA exemptions were criticized by Sens. William Roth, R-Del., and John Glenn, D-Ohio, senior members of the Senate Government Affairs committee, who argued that the proposed exemptions provided no safeguards against contracting or personnel abuses. The Senate committee, and its House counterpart, also worry that PBO exemptions will chip away at governmentwide laws under their jurisdiction.

Reaction to the proposed exemptions for the Patent and Trademark Office has been harsh. In Congressional hearings last March, leaders of the agency's three unions opposed HR 2533, largely because of their concern over the proposal to exempt the agency from civil service legislation. Ronald Stern, president of the Patent Office Professional Association, said that the bill would "eliminate the underpinnings of the civil service system" and allow managers to "carry out their personal prejudices."

Although senior officials at OMB and OPM support the PBO plan, some staff in those agencies share concerns about giving PBOs broad legislative exemptions. They argue that some PBO candidates have neither made the case for broad exemptions, nor considered whether existing regulatory flexibility might meet their needs. One OMB official says he's "convinced that a large proportion of what people would like to do could be done with presently available flexibilities."

Central management agencies are now making an effort to show PBO candidates how much freedom they've already got under existing law. For example, OPM distributed a memo in March that outlined existing flexibilities, such as the right to increase pay under special circumstances. It held a workshop on flexibilities for PBO candidates in May. Other agencies are taking similar steps.

Where PBO candidates do propose legislation, it's unlikely to include broad exemptions like those given to FAA and proposed for PTO. Instead, two alternative approaches are being considered. In the personnel area, PBOs would be allowed to launch demonstration projects like those authorized in the 1978 Civil Service Reform Act. Some of the rules that restrict demonstration projects in the 1978 Act would be waived, but OPM's power to approve and supervise projects would be maintained, and unionized employees would be protected against dramatic changes.

The Office of Federal Procurement Policy is taking a slightly different approach. It has developed a menu of provisions that provide relief from existing procurement rules. Many of these provisions are based on proposals for governmentwide procurement reform that have not yet been adopted by Congress. PBO candidates would select those menu items that suit their needs.

Getting Results

Under the PBO plan, the trade-off for increased freedom is tougher accountability for results. The plan builds on the 1993 Government Performance and Results Act (GPRA), which requires agencies to develop annual performance plans and performance reports that establish and report on specific, measurable targets. Targets would be negotiated by each PBO and its parent department at the start of the year, and a large part of each PBO executive's pay would be tied to meeting those targets.

Target-setting sometimes fails when managers try to establish a definitive list of outcome measures for their programs. The British have taken a narrower, more practical approach. Annual performance agreements focus on a small number of measures that highlight aspects of agency work that need attention, such as processing time for benefit claims, or accuracy of agency decisions. Targets also focus on productivity and unit costs, or specify budget savings to be achieved in a year. Measures often change from year to year as agencies improve their understanding of what matters and their ability to collect performance information.

However, British experience has shown that making and enforcing performance agreements is not always easy. A 1994 study of the British reforms by French civil servant Sylvie Trosa endorsed the new system but reported that parent departments were sometimes frustrated by their inability to judge whether targets were tough enough. As a result, Trosa, observed, "most departments take the view that the game is to be more demanding than agencies."

There have been other problems, too. Trosa found that in the early years, departments leaned more on what could be measured than what really mattered. That meant more emphasis on financial targets than on service quality, if the agency did not have good systems for assessing service. Trosa also cited the example of a regulatory agency that was encouraged to focus on the number of enforcement actions, which could be easily counted, rather than industry compliance with the goals of the law.

GPRA's phase one pilot projects, which focus on the development of performance plans, have produced similar results. An OMB study completed last fall said that GPRA's pilot projects showed high-quality performance plans could be produced across government. But the study found "a significant number of pilot projects encountered substantial problems." It noted agency difficulties in defining measurable targets, a focus on measures that were unrelated to program goals, and difficulties in obtaining data needed to assess performance.

The administration wants to give teeth to the idea of accountability for performance by changing the way PBO executives are hired and paid. Its proposal for the patent office provides that the new chief executive officer would be hired on a six-year contract and could only be fired for failing to live up to the terms of the yearly performance agreement. The CEO's base pay would be $148,000, but an annual bonus of up to $148,000 could be awarded for hitting performance targets.

OMB's Koskinen calls this new way of hiring and paying executives "the main argument in favor of the PBO model." He notes that service delivery functions "often are run by political appointees who turn over in less than two years and often are more interested in policy formulation rather than management." Career executives, on the other hand, are "generally very good," but can't be dismissed if performance is "not bad but not great."

However, the PBO hiring and salary arrangements might be tough to implement. It isn't clear that Congress will tolerate a total salary package of almost $300,000. Advocates argue that the pay is lower than that of executives heading businesses of similar size. In addition, questions have been raised about the constitutionality of the administration's proposals for appointment and removal of the patent office chief executive. NPR's Kamensky says the Justice Department has approved the approach taken in the administration's PTO bill.

The bonus plan would also have to be handled carefully. The proposed 100 percent bonus is much larger than that paid to the heads of Britain's executive agencies. Unless the system for assessing performance and deciding on bonuses is perceived to be fair, and there are mechanisms for quickly resolving disagreements about bonuses, the plan could meet opposition. Similar concerns arose after the introduction of performance-based bonuses for Senior Executive Service members in 1978.

New Zealand's government, which uses performance-based pay for its department heads, tries to manage this problem by having a third party-its State Services Commission-make decisions about bonuses. Jonathan Boston, a professor of public policy at Victoria University at Wellington, has studied the New Zealand system, and says most executives like the performance-based approach, because it gives them clear annual objectives. Cabinet ministers, on the other hand, sometimes express frustration about the rigidity of the system and the time consumed in negotiating agreements.

Who's in Charge?

The administration's proposal to establish the patent office as a PBO was widely panned during Congressional hearings last September. All three of the agency's major user groups-the American Bar Association, the American Intellectual Property Law Association and the Intellectual Property Owners-opposed the proposal. User groups preferred an earlier bill introduced by Rep. Carlos Moorhead, R-Calif., that would establish the office as a conventional government corporation. Underlying the debate over the PBO proposal was a tussle for control. User groups believed that the administration's proposal would have given the Commerce Department too much power to meddle in patent office activities. Moorhead's bill would have given the office more independence. Its CEO would be confirmed by the Senate, and would have to consult with an advisory board of user representatives mostly appointed by Congress.

In May, the subcommittee that oversees the patent office decided to ignore the administration's proposal and report out Moorhead's bill. Administration officials protested, arguing the Moorhead measure undermined the Administration's ability to make policy in the intellectual property field. "The advantage of the PBO concept," OMB's Koskinen argues, "is that there is no doubt that the operating entity is still part of the department."

In June, the Administration warned that the Moorhead bill would be vetoed if changes weren't made to strengthen the Commerce department's power over the patent office. One amendment to clarify Commerce's role was made in a Judiciary Committee mark-up of the bill. Another amendment requiring yearly performance agreements between the CEO and the department will be introduced on the House floor.

The PBO plan is consistent with earlier NPR recommendations that also strengthened the administration's control over administrative activities. In fact, the history goes back further than that. Executive branch reformers have been using the British civil service as their model for most of this century. The British model, advocates argue, gives top managers more room to use resources efficiently and engage in long-term planning. The price, critics complain, is reduced legislative oversight and control over administrative activities.

The PBO plan may be modified to respond to these concerns. For example, Congressional committees might be given an opportunity to discuss annual performance agreements before they're finalized with PBO executives. None of the 132 independent agencies will be candidates.

Advocates of the PBO plan will also attack the argument that it diminishes Congressional control over departments and agencies. NPR's Kamensky says the administration's goal has not been to diminish oversight but to direct attention away from administrative details toward outputs and outcomes. Furthermore, proponents argue, the PBO approach will provide the information-in terms of performance targets and reports-that is needed to exercise effective control over PBOs.

Good Fences

The biggest difficulty confronting the UK's Next Steps agencies gets little public attention. It's the problem of setting and enforcing boundaries between parent departments and the executive agencies. The 1994 Trosa report called this "the most difficult challenge facing the Next Steps reform."

Executive agencies complain that parent departments interfere too much in agency activities, or refuse to give agencies the freedom they need to achieve their targets. In some cases, agencies haven't used the flexibility they've won because they fear angering the parent department and provoking more intervention.

Parent departments sometimes have good reasons for wanting to exercise tight control over executive agencies. They cite their obligation to coordinate policy or service delivery between agencies, worries about erosion of staff mobility within the department, or their responsibility to ensure that agencies use resources prudently and efficiently. Studies of similar reforms in other governments have found that parent departments sometimes worry that agencies will become loose cannons whose activities might embarrass the department's political head.

The problem of managing the department-agency relationship isn't a new one. A 1991 study of the British reforms suggested part of the difficulty might be a lack of clarity among parent departments about their role in the new system. Trosa suggested that interfering with agencies might be a natural response of parent departments threatened with substantial budget cuts and seeking to rationalize their own existence. The 1991 study proposed that senior civil servants within departments should be appointed to act as impartial arbitrators between departments and agencies. Trosa made the same recommendation.

The public also may pressure departments to intervene. The British Prison Service, set up as an executive agency in 1993, found this out last year. The service attracted controversy after two escape attempts by high-risk prisoners. There has also been debate about whether the service is too soft on dangerous offenders. Labor Party critics pressured the minister responsible for the Prison Service, Michael Howard, to tighten the reins on the agency.

Howard had a dilemma. When he argued that the problems were operational, and that agency managers should be given room to fix them, opposition party legislators criticized him for "passing the buck." When Howard tried to exercise tighter control, however, agency managers complained that he was violating the terms of the performance agreement. After months of controversy, Howard eventually fired the agency head, Derek Lewis, last October.

Lewis quickly announced he was suing Howard for wrongful dismissal. He noted that the Prison Service achieved all of its performance targets in 1995, and complained that Howard's interference undermined his ability to get work done. In March, the British government, wanting to avoid a lengthy public trial, announced it would not contest Lewis' lawsuit.

There are more than a hundred British CEO appointments, and Lewis is the first to sue for wrongful dismissal. The Prison Service was also a bad candidate for executive agency status, because of its high public profile and lack of consensus about the work it does. In March, NPR said it won't recommend PBO candidates that don't have "broad-based support from key stakeholders."

Rules or Culture?

Some critics are likely to argue that the PBO plan focuses too much on the rules that constrain line managers and not enough on the real barriers to change: managers with weak leadership skills and organizational cultures that kill initiative and innovation. In last summer's debate on the FAA exemptions, for example, Sen. William Roth, R-Del., argued that the organization's troubles were "a result of poor management," not legal constraints. Critics also point to a recent General Accounting Office examination of NPR's "reinvention labs" that suggests the importance of waivers as a prerequisite for reform is overstated.

PBO advocates don't deny that reform requires more than loosening the rules. Heads of the PBOs will have to work hard to retrain workers, build a service-oriented culture and encourage initiative. But proponents argue that attempts at cultural change will fail if rules make it impossible for employees to be innovative or responsive to customer demands.

Supporters of the PBO plan also emphasize the essential second element: a heavy emphasis on results. They argue that the PBO approach will force organizations to get better at specifying goals, measuring performance, and incorporating performance data into day-to-day decision making For people outside the executive branch-whether in Congress or the media- the PBO plan could make it easier to decide whether agencies are getting the job done.