Setting Goals for Government: Revising GPRA

Here’s a re-posting of a piece I co-authored with Jonathan Breul, Robert Shea, and Jitinder Kohli on the Center for American Progress’s “Doing What Works” website. It’s significant in that it is by senior staffers who worked on government performance issues from the Reagan, Clinton, and both Bush administrations. It explains why the Senate should embrace a bill to revise GPRA to improve government performance.

The Senate returned this week to an exceedingly busy legislative agenda, but we urge its members to take up a bill that could transform the way government works by requiring agencies to set priority goals and achieve real-world results. The legislation, which has yet to be introduced in the Senate, should be modeled on the bill introduced by Rep. Henry Cuellar (D-TX), which was passed unanimously by the House in June. This legislation requires each federal department and agency to set high-priority goals they intend to achieve for the American people. Timely enactment by the Senate and President Obama’s signature before year’s end would demonstrate the federal government is clearly focused on improving its performance and winning back the confidence of the American people.

The legislation would require all departments and agencies not only to set high-priority goals but also develop a plan to accomplish each priority goal, monitor progress quarterly toward the goal, and regularly adjust the plan as they go along so that the goals are more likely to be met. This closely follows an approach that the Center called for recently to help policymakers build confidence in government. Setting agency goals that target real-world results for the American people will force leaders to prioritize initiatives that are the most effective at achieving these results. The fiscal context facing all agencies amplifies the importance of adopting this approach.

It is also what the American people want. A recent poll by the Center found that 83 percent of Americans believe requiring every agency to set clear goals would make government work better. The legislation passed by the House and now before the Senate would enact a strategy to improve government that the American people strongly favor.

Senators on the Homeland Security and Governmental Affairs Committee should quickly pass legislation building on the House version, and even in the final days of a busy session, they should get the bill to the president’s desk. They’ve got the support not only of a unanimous House, but also of Senator Mark Warner (D-VA), who, as governor of Virginia, oversaw reforms that set goals at the state level. He now chairs the Senate Budget Committee’s Taskforce on Government Performance.

In short, the stars are aligning in support of this important measure. But a key question facing the Senate will be how far-reaching it wants the legislation to be. We believe there are three main options:

** Sharpen legislation passed by the House so that it is more effective.

** Be more expansive and address the connection with the Government Performance and Results Act of 1993 (GPRA) so that GPRA works well alongside the new requirements imposed by this legislation, or perhaps go further and undertake more fundamental reform of the GPRA to make it more useful and used.

** Ask President Obama to set a small number of top-level goals that cut across government departments.

These options are not mutually exclusive. Indeed, we believe that the bill would be strongest if the Senate decided to enact all of these options at the same time and thus build a new performance management framework for the federal government.

Let’s consider each option in more detail.

Sharpen the House bill

This would entail a focus on agency, not program, goals. The Center earlier this year examined the key ingredients needed to make this bill effective. The House embraced many of these suggestions as the bill moved to the Senate, but some aspects of the bill could be sharpened. For instance, the House bill was amended to require goals to be about outcomes that “address the needs of the American people.” But the new requirement applies to program-level—not agency-level—high-priority goals. The Senate should insist that it apply at the agency level and only for priority goals, not for every goal for every program.

Limit the number of high-priority goals
There should be a limit of perhaps five high-priority goals per major agency or an overall ceiling of 100 across all major agencies because they are meant to reflect priorities. Remember, if you’re measuring everything, you’re measuring nothing. Each goal should have a responsible official and a clearly defined level of improvement that everyone agrees represents accomplishment of the goal.

Consultation with Congress is imperative
The House bill makes clear that Congress should be consulted before agency goals are adopted. This is essential, yet the Senate should also set out a timetable for the process. We believe agencies should issue a set of draft goals by October in the year a president takes office and seek views from the public, stakeholders, and Congress before finalizing goals by February of the following year.

Independent review is essential
One risk of the House bill is that poor goals will be set. Goals might be unrelated to outcomes, or insufficiently demanding, or the data used to measure progress toward the goal might be of poor quality. To guard against this risk, we think the bill should ask that the Government Accountability Office or the Congressional Budget Office take on the role of commenting on draft goals against these criteria.

Limit goal revisions
At the first sign of failure, agencies may well seek to revisit goals by lengthening the timeframe over which the goal should be achieved or by decreasing the targeted level of improvement. There will be instances where such changes are merited, but many more where agencies seek them just to make their lives easier. The Senate should make it clear that changes such as these can only be made in exceptional circumstances, and require an agency to publish in advance the rationale for the change, and seek the views of the public and Congress before revising the goal.

Address the connection with the Government Performance and Results Act

The Senate could go further and address the connection with the Government Performance and Results Act. The House bill is merely an addition to the existing GPRA requirements, requiring agencies to set agency-wide priority goals and evaluate programs against those goals. But the GPRA already requires each agency to adopt a strategic plan, publish an annual performance plan, and prepare an annual performance report. In addition, the GPRA also requires that in these documents, every program of every agency must set its own goals and indicators, and measure performance against those. The House legislation further complicates this by amending the GPRA to require each program to explain the outcome that it is trying to address and the specific impact that the program will have on that outcome.

By adding to the requirements of the GPRA, the new bill potentially exacerbates rather than solves one of the main weaknesses of the GPRA—complex, compliance-focused measuring and reporting. Current performance-management practices in the federal government are often too complex and confusing to be useful for managing or overseeing agency operations. The GPRA already ensures that agencies produce lots of information, but few decision-makers make use of the information. To some extent, that’s because of the sheer volume of information produced. This is a recipe for confusion and does little to improve government performance.

Exempt agencies from double reporting
One way the Senate can fix this problem is to exempt agencies from double reporting under the GPRA and the new bill. The Senate could legislate that if a program can demonstrate how it is contributing to the overall priority goals of the agency then it does not need its own goals and indicators too. The Senate should consider exempting such programs from the old GPRA requirements. That would mean that for an individual program, either the new approach would apply or the old one—but not both at the same time.

Programs that do not contribute to priority goals could report voluntarily
Exemption from double reporting would leave GPRA requirements largely intact as many programs will not contribute to priority goals. There is a strong case for going further as few in Washington believe that current GPRA-reporting requirements help improve performance. It might be more sensible, then, for the Senate to reform the GPRA so that it no longer requires agencies to produce voluminous reports. The new legislation could limit reporting requirements to high-priority goals—requiring that agencies describe which agency programs contribute to a high-priority goal and to what extent they will help accomplish the goal.

In this way, those programs that do not contribute significantly to an agencywide or governmentwide priority goal could be encouraged (rather than required) to set outcome-based goals for what they will achieve. Many agencies will likely choose to do so voluntarily since this will help them manage performance internally and account to the American public for the money that they spend. But in cases where it is hard to set outcome-based measures for the success of such programs, they may choose not to do so rather than setting poor-quality ones.

Ask the president to set a small set of top-level goals that cut across government departments

The House bill requires agencies to set priority goals, but there is also a strong argument for setting a smaller number of government-wide, top-level outcome goals in addition to agency-level goals. The most important issues facing the nation are often ones that cut across agencies. Responsibility for reducing unemployment, for example, cannot belong to the Departments of Labor, Treasury, and Commerce; and the Small Business Administration alone. Instead, creating jobs is something that they and other agencies should contribute toward, with one agency taking the lead on co-coordination.

Similarly, reducing childhood obesity is not an issue that the Departments of Agriculture, Health and Human Services, and Education can address separately. In fact, the recent Taskforce on Childhood Obesity brought together 12 departments to develop a plan to tackle the issue.

The Senate bill could require the administration to set a small number of government-wide goals that define the most important outcomes that the administration wants to accomplish for the American people. They should be set by the executive branch within six months of a president taking office. There could be 5 to 20 cross-cutting, top-level goals of this nature so that they are clear to the American people.

Finally, once government-wide goals are set, there also needs to be a process to ensure that they are accomplished. The Senate should require each goal to have a senior official responsible for ensuring that all departments and agencies work together to accomplish the goal. The president should personally ensure regular meetings to check on progress toward these goals. This would build on the successful approach adopted by cities such as New York and Baltimore, states such as Washington and Maryland, and countries such as the United Kingdom. In these cases, the mayors, governors, or prime ministers all personally oversaw efforts to accomplish challenging goals with extraordinary results.

Seize the opportunity to transform government performance

In 1993, when Congress passed the Government Performance and Results Act, it believed there needed to be a much sharper focus on whether individual programs were working. This approach has certainly increased the amount of available performance information. But it has had less impact on the culture of government. Agencies or programs still struggle to define what outcomes their work will achieve for the American people. And the American people struggle to understand how the work of agencies will make their life better.

The House legislation offers an opportunity to build a new, more effective performance-management framework for the federal government. We encourage the Senate to bring forward legislation of its own that builds on the House bill.

All of the changes we argue for in this column would enhance the final legislation the president would sign. Indeed, with these changes the Senate could build a federal performance-management system that finally gives Congress the practical information it needs to focus limited resources on activities proven to produce the greatest results for the American people.

That is why it is so important that the Senate should devote some of its scarce time to this very important issue. Results are what matter to the American people.

* * * * * * *
Jonathan D. Breul worked in the Office of Management and Budget to reform government-wide general management policies from 1980 to 2002. He helped develop the President’s Management Agenda, the President’s Management Council, and championed efforts to integrate performance information with the budget process. He led the development and government-wide implementation of the Government Performance and Results Act. Currently he is executive director of the IBM Center for the Business of Government.

John M. Kamensky served for eight years as deputy director of Vice President Gore’s National Partnership for Reinventing Government. Before that, he worked at the Government Accountability Office playing a key role in the development of the Government Performance and Results Act. He is a senior fellow with the IBM Center for the Business of Government.

Jitinder Kohli is a Senior Fellow at the Center for American Progress focusing on government efficiency and regulatory reform. Prior to that, he spent 15 years in the British government. Between 2005 and 2009, he led the UK’s work to reduce the administrative burden of regulation.

Robert J. Shea led the Performance Improvement Initiative at the Office of Management and Budget during the Bush administration. Prior to that, he was counsel to the U.S. Senate Committee on Governmental Affairs. He is now a partner in Grant Thornton’s Global Public Sector and has extensively published on government performance issues.

This piece is part of CAP’s Doing What Works project.