Discover more from The Bulwark
We Used to Want Government to Work Well—And We Still Can
The surprisingly long history of the quest to make bureaucracy more effective and efficient.
WHEN YOU THINK BACK to the most consequential policy measures of the Clinton era—the North American Free Trade Agreement, welfare reform, and a handful of others—one that probably does not leap so readily to mind is the Government Performance and Results Act of 1993, a law intended to improve the functioning of the federal bureaucracy. But GPRA, despite its relative obscurity and the fact that it never quite delivered the transformation that its authors sought, was groundbreaking in its own way and deserves to be remembered. With the law’s thirtieth anniversary approaching later this summer, now is an appropriate time to look back on GPRA’s origins, take stock of how well it did what it supposed to, and consider what might come next.
GPRA was Congress’s response to the widespread sense that many federal agencies suffered from organizational drift. The intention was to force all of them to bring more focus to their long-term missions and to take seriously the need to establish clear metrics for periodically assessing their progress.
The sense that the government would benefit from a performance breakthrough has not gone away even with GPRA in place. With fifteen cabinet departments, scores of additional agencies, and more than 10 million direct and indirect employees, the federal enterprise is vast, and yet it also struggles to address the country’s proliferating array of pressing social challenges. A national government capable of delivering more with less would ease some of the political and economic tensions that have been rising in recent years.
It is also not difficult to imagine what areas might be most ripe for improvement. The private sector has gone through multiple stages of transformation in the last three decades as information technology upgrades opened up possibilities for delivering services to customers in new ways and at less cost. Many federal services and processes might be similarly streamlined.
The challenge is to find a formula that incentivizes federal agencies to voluntarily implement efficiency-enhancing reforms even when doing so risks a reduced appropriation in the future. GPRA was an important milestone for the performance-focused movement, but it did not provide a convincing response to the default position of the bureaucracy, which is protection of the status quo.
Speaking of the status quo—Lincoln said, “The dogmas of the quiet past, are inadequate to the stormy present. . . . As our case is new, so we must think anew, and act anew.” That’s what we’re doing here at The Bulwark. Join us.
Managing Entities Without Bottom Lines
ALTHOUGH IT MIGHT SEEM that attempts to make the federal bureaucracy more efficient and responsive are a fairly recent development in our national political life, the performance movement in public administration in fact goes back more than a century. The fundamental problem it seeks to address is that governmental agencies lack the ready accountability that profitability imposes on private businesses. The first initiatives aimed at providing alternative sources of discipline were implemented by state and local governments in the Progressive era.
After World War II, federal administrations of both parties intermittently prioritized improved management of the national government’s expanding bureaucracy even though taking on the challenge never generated much public enthusiasm or interest. A consistent concern has been a federal budget process that places too much emphasis on resource provision and not enough on delivering tangible results for what is spent. The absence of meaningful performance data for many activities also makes it challenging to enforce accountability.
Calls for improvement have come most often from within the executive branch itself and not Congress, which is understandable given their differing constitutional roles. Among the more notable initiatives have been: Defense Secretary Robert McNamara’s planning, programming, and budgeting (PPB) system for the military in the Kennedy and Johnson years; President Carter’s zero-based budgeting (ZBB) formulation; the many privatization recommendations of the Reagan-era Private Sector Survey on Cost Control (the Grace Commission); and the Program Assessment Rating Tool (PART) developed and implemented during the George W. Bush years.
As the Clinton team made its first moves in early 1993, key officials viewed restoring confidence in the competency of the federal bureaucracy as crucial to their plans. An influential 1992 book, Reinventing Government by David Osborne and Ted Gaebler, reflected the ambitions of the incoming administration and caught the attention of Vice President Al Gore. The management initiative Gore led over the ensuing eight years, known as the National Performance Review, encouraged experimentation with a variety of management strategies. Osborne served as a top adviser.
At the same time, the government’s primary performance watchdog, the Government Accountability Office (then known as the General Accounting Office), was getting traction in Congress with a recommendation for a governmentwide performance accountability system. GAO had long argued that an emphasis on pursuing measurable results would help improve the productivity of agencies stuck in a cycle of mediocrity. The GAO-inspired bills introduced in the House and Senate in the years immediately preceding 1993 stalled at various stages of consideration but were nonetheless instrumental in building momentum for the initiative.
As the Clinton administration took office, the preparatory work of the previous years paid off, with the final version of GPRA passing easily through the House and Senate by way of consent agreements (which imply unanimity). President Clinton signed it into law in August 1993.
The Limits of Reporting-Focused Accountability
GPRA IS IMPORTANT BECAUSE it forced federal agencies to institutionalize some of the most basic requirements of sound organizational management. It put into permanent law a standardized and governmentwide performance reporting system with three components: (1) a periodically updated agency strategic plan which lists clear, overarching organizational objectives; (2) an annual goal-setting report focused on key metrics of performance; and (3) transparent disclosure of actual results.
As an example, the Social Security Administration’s current goals include shortening hearing wait times for disability applicants, reducing erroneous payments in the Supplement Security Income (SSI) program, and shortening wait times for calls to the agency’s toll-free call center.
The expectation was that requiring this information to be fully transparent and publicly accessible—there is now a federal website dedicated to publishing federal performance information—would provide strong incentives for agency leaders to set ambitious goals, and to meet them.
The reality has been somewhat different.
While GPRA is credited with forcing agencies to grapple with performance questions in ways that previously could be avoided altogether, the immense volume of reports, performance measures, and corrective plans has proven to be unwieldy, with the important often lost in a never-ending stream of data, much of which is not useful.
Further, it has proven exceptionally difficult to use GPRA to incentivize real change in operational practices, for a few reasons.
First, there has always been some distance between the accountability that performance advocates imagine to be possible and the reality of the public sector. In the private sector, if a corporate division consistently falls short of its objectives, you would expect there eventually to be decisive consequences—maybe firings or hirings or a reorganization or shuttering. By analogy, you might imagine that if a federal agency consistently falls short of its objectives, it might be defunded, or at least have its budget cut, with its the resources then redirected to agencies and programs that have demonstrated a capacity for delivering better value to taxpayers.
That almost never happens. A major reason is that most federal functions—think enforcement of food safety, for instance—are strongly supported by the public. If a government agency is falling in its mission, it is sure to make the argument that it fell short in part because of inadequate resources. More often than not, politicians will agree—and then approve not budget cuts but an expanded budget for the agency in question.
Second, there is no penalty for low expectations. Experienced federal managers long ago figured out there was no real reward for setting ambitious goals, and possibly much to lose. The result is a system that is filled with targeted objectives that mostly reflect business-as-usual and not major performance improvements.
Third, again, Congress has been less interested in enforcing performance accountability than the executive branch. Most agencies and programs have champions in Congress, for many possible reasons, including the locations of key facilities and possibly a high concentration of program beneficiaries in certain areas. The presence of these champions in key decision-making positions makes it difficult for others to push for accountability. The appropriations process does not currently exhibit any real tendency to distinguish between high- and low-performing programs and agencies.
The PART Scorecard
GPRA’S CHALLENGES, ALONG WITH the many complaints from the agencies about its burdens, were well understood as early as 2001 as the George W. Bush administration got underway. Incoming officials wanted to maintain a strong focus on performance improvement but with a streamlined process. The result was the development of the Program Assessment Rating Tool (PART), which did not displace GPRA but was used more rigorously within the executive branch during the period 2002 to 2009.
The PART was a question-and-answer-focused evaluation device that translated a series of yes-no responses and also measures on program results into a performance score for each relevant subagency or programmatic activity of the federal government. The idea was to provide executive branch leadership and Congress with a digestible and fair system of program assessment that would allow ready identification of high and low performers. The Bush administration incorporated the PART process into its internal budget deliberations starting with the fiscal year 2004 budget. (Full disclosure: I was at the time an associate director of the administration’s Office of Management and Budget.)
While the PART became an accepted practice in the Bush administration because of high-level support, it also experienced pushback from the permanent civil service and from Congress. Agency staff saw it as redundant of GPRA and also a thinly veiled justification for budget cutting. Many in the House and Senate saw it as an extralegal mechanism, not authorized by law, for shifting power in the appropriations process away from Congress.
Resistance to the PART from the appropriations committees was particularly damaging to the administration’s plans as Congress’s cooperation was essential if the tool were ever to influence budgetary decisions. But there was never an indication that buy-in from Congress was forthcoming.
When the Obama administration came into office, the PART was phased out to make room for a renewed focus on strategic goals and reliance on the GPRA framework.
BY 2009, IT WAS CLEAR THAT, while GPRA had institutionalized performance reporting, it was not leading to major breakthroughs in productivity and better services, partly because it had been mostly downplayed during the preceding eight years of the Bush administration, which had invested its energy in the PART. Key Obama officials wanted to work with their allies in Congress to revitalize GPRA by updating it, which occurred with the enactment of the GPRA Modernization Act in January 2011.
The updated GPRA made several changes to the original law. It required the government to establish a priority goal system, to focus greater management attention on the key metrics that really matter to overall agency performance. In the years after GPRA was enacted in 1993, most agencies identified scores of measures that tended to dilute the importance of all of them.
The revised GPRA also called on the government to establish more cross-agency strategic objectives, to help focus the entire federal enterprise on crucial national goals, and required federal offices to track their key goals on a quarterly basis to ensure the data was more relevant for real-time management decisions.
Finally, the law reworked the accountability system and required the Office of Management and Budget to report to Congress on the remedial steps agencies were taking when they repeatedly missed performance targets. Among the possibilities is a recommendation from OMB to Congress to terminate programs that provide poor value for what is spent on them.
Nothing says “value for money” like a Bulwark+ subscription. Join for free, or subscribe to unlock all the special member benefits.
AS WE’VE SEEN, a big part of why it’s so difficult to move the public sector toward delivering services less expensively is that it is difficult to incentivize bottom-up reform. GPRA, PART, and GPRA 2.0 have all been top-down endeavors; Congress and executive branch leaders have used them to impose new performance-focused reporting requirements on agencies. For the most part, the bureaucracies have complied, but often reluctantly. These weren’t their ideas.
Real change will require flipping the script and allowing the agencies to take the initiative because it is the officials with extensive knowledge of existing practices who are most likely to have ideas for reducing costs. They will only be enticed into disclosing what they know if there is a real payoff awaiting, for themselves and their agencies, if the plans that follow produce the desired results.
Another initiative from the Clinton years might serve as a useful starting point for devising and testing a new approach to performance improvement. As part of the National Performance Review, the administration borrowed an idea from the Thatcher era in the United Kingdom—performance-based organizations (PBOs)—and modified it for the U.S. context. The concept was to offer operations-focused agencies a deal: they would receive enhanced autonomy to manage their affairs with less interference (including with respect to compensation) if they committed to achieving concrete performance goals.
Initially, there was some interest in the idea in Congress, which led to the approval of a PBO structure for the Department of Education office responsible for managing student aid. That agency was able to turn around its performance when given more room to manage its affairs as a standalone entity. However, there is also a concern in Congress that giving agencies wide latitude to pay their employees more diminishes their accountability to Congress. After the Clinton administration, interest in the PBO concept waned.
It is time to bring PBOs back. Designed properly, they can be powerful because they are agency driven. What was missing in the 1990s was a fairer distribution of the benefits. As envisioned by the Clinton administration, the financial benefits of improved productivity would accrue almost entirely to the agency itself, in the form of higher pay and bonuses and more investment in future operations (separate federal laws authorize bonuses for high-ranking federal managers and other staff outside of the PBO context, but PBOs were drawn up to allow even greater flexibility, including with respect to basic pay levels). A new version might be more attractive to Congress if the benefits were shared with taxpayers in addition to incentivizing federal managers to deliver better results. For instance, a PBO might get to spend some of the savings from improving efficiency only if it reduced its costs below an agreed-upon baseline and returned half of the difference to taxpayers by way of payments back to the Treasury.
A further advantage of PBOs is they require active congressional participation. With GPRA and GPRA 2.0, Congress has been mostly a passive observer of agency compliance. Federal offices write the required lengthy reports, but Congress is under no obligation to read them or alter its decision-making based on what is disclosed.
With PBOs, Congress would have to explicitly authorize new incentive structures for the agencies, and thus also decide how they should be designed and administered. It would bring Congress into the conversation about how to more effectively manage the federal bureaucracy. To assuage concerns about accountability, Congress could limit the conferring of PBO designations to five-year increments, subject to oversight processes designed to quickly consider requests for extensions.
BETTER MANAGEMENT OF THE IMMENSE federal enterprise is not a topic that elected leaders are eager to take on because the political dividend is likely small or nonexistent. Improving the performance of complex public sector operations is often a messy affair, with a mix of progress and setbacks along the way. Officials usually want more certainty that their efforts will lead to positive reviews from voters.
Still, even with these challenges, it is important to prioritize high performance in the public sector. There are certain roles that private enterprises will never take on. If the public agencies that are handed the responsibility for these essential tasks also drop the ball, it is dispiriting for voters because they have nowhere else to turn. This is one reason trust in institutions continues to erode.
If, on the other hand, the federal government were seen as making steady progress as a “customer”-friendly enterprise while also imposing less of a burden on taxpayers, the cumulative effect over time could be healthy in a society that is suffering from an excess of cynicism.