Few issues could be of greater importance to the work of these committees than the implementation of the Government Performance and Results Act of 1993 (GPRA). It is a great pleasure, therefore, to appear before you to discuss the critical stakes for Congress in effective implementation of this act.
I am a Professor of Public Affairs and Political Science at the University of Wisconsin-Madison. I am also a Nonresident Senior Fellow in the Brookings Institution’s Center for Public Management, where we have been closely following GPRA and the broader issues of improving government performance.
What I want to share with you today are ten reasons why GPRA can help Congress solve problems that must be solved. My argument, in summary, is this:
Citizens and public officials alike are all too correct in criticizing the way the federal government works. Performance is not what it could be–and the performance measurement required by the act is the keystone for solving this problem. But Congress is critical to making performance-based management work. Without strong congressional leadership, GPRA’s impact will fade away. For members of Congress to play that leadership role, GPRA will have to provide them with answers to questions they need to solve. GPRA can in fact do just that: It can provide the missing link between authorizations and appropriations, on the one hand, the worries about government performance on the other.
The ten points are these:
1. We must tackle both the performance and budget deficits.
2. Performance measurement is the keystone to reducing both deficits.
3. Performance measurement provides Congress with critical information about agencies’ strategic decisions
4. Strategic plans provide a road map to achieving results.
5. Performance measurement connects plans with results.
6. Performance management can improve the authorization process.
7. Performance measurement can improve the appropriations process.
8. Performance measurement is no magic bullet — but it helps congress do what has to be done.
9. Performance measurement can transform the president’s budget submission to Congress.
10. Performance measurement can vastly improve congressional policymaking.
Let me explore these points in turn.
1. We Must Tackle Both the Performance and Budget Deficits
Too often, debate on cutting the federal deficit proceeds on the assumption that, should we ever truly succeed in eliminating the deficit, the nation’s problems would end. Eliminating the deficit would demonstrate great political will, and that would be a huge accomplishment. Citizens worry that government doesn’t work and can’t make tough choices. A serious deficit reduction plan would show that government solve such problems.
Even balancing the budget, however, would not solve all the critical questions. Eliminating the deficit, whether by cutting programs or increasing taxes, would necessarily not ensure that government would work any better. Indeed, some programs might work even worse. Many of the downsizing and budget cutting decisions have been made with a blunt instrument. Government offices have often come out of the process ill-configured, with the wrong collections of people and technology to promote efficiency. (That has too often been the result of downsizing in the private sector as well.) Inefficient offices could frustrate citizens doing business with government; they could allow fraud, waste, and abuse to grow; and they could generate fresh news reports about a government that can’t shoot straight.
Citizens are not likely to react warmly if, after having been promised a government that works better and costs less, they get a government that works worse and costs more. Applicants for social security benefits expect quick and friendly service, along with predictable and accurate checks. Midwest residents want timely tornado warnings, while coastal residents need good hurricane predictions. Workers expect their pensions to be safe; flyers, the air traffic control system to guide them safely to the ground; citizens, their drinking water to be free of toxins. The symbol of a balanced budget will have little meaning if citizens are angry that the services on which they rely are poor and unreliable.
That is not an inevitable outcome. But it is a likely result unless government’s policy makers realize that shrinking government’s size is not one problem but two: first, eliminating what can and should be cut; and second, ensuring that what is left—what policy makers determine is the core that government can and should manage—works well. So far, public debate has focused on the first problem. Sooner or later, we will have to turn to the second. And unless we solve it, public anger at government, its institutions, and its elected officials is likely to grow even bigger.
I have argued elsewhere that American government has not one deficit but two: a budget deficit, and a performance deficit. Unless we solve both deficits, citizens rightly will be unforgiving.
GPRA frames a coherent attack on both the budget and performance deficits by allowing a serious discussion, for the first time in American government, on what we get for the money we spend. It provides members of Congress a way to bridge the gap between the budget and performance deficits.
2. Performance Measurement is the Keystone to Reducing Both Deficits
We are not alone in struggling with the twin deficits. Some nations have been working for well over a decade on cutting both deficits, while the American federal government has been single-mindedly focused on the budget deficit. Other nations have had greater success in reducing both deficits than the United States has. And the most successful nations have build their efforts on performance measurement.
Although the private sector can provide interesting insights for performance, the really useful models on these questions are countries like Australia, France, New Zealand, and Sweden, and the United Kingdom. The drive to shrink government has been surprisingly global. Where it has been most successful, the government reduction movement has been coupled with tactics to improve government performance. In Australia, the focus has been on “letting managers manage” by working to remove barriers to energetic administration. In New Zealand and the United Kingdom, the focus has been on “making managers manage” by introducing market competition into government services. But in all of these cases, performance measurement has been at the core.
Performance measurement in these countries has been important for two reasons. First, performance measurement has helped government policy makers move from a focus on inputs—how much they spend on programs—to a focus on outputs—what results they get from the money they spend. The American reform effort has concentrated just on the first half of this equation. It is little wonder that it has proven so frustrating, because the input focus (how much money to spend) has proceeded independently of what really matters, what the money buys. With this disconnection, it is hard to ask the important questions or to provide answers that satisfy anyone for long.
Second, performance measurement has helped government policy makers move from process—how decisions get made—to results. For a decade, the American federal government has struggled to devise procedural solutions for tough substantive problems. Elected officials have tried automatic deficit reduction tactics like Gramm-Rudman and have proposed spending ceilings like the balanced budget amendment. No process, however, can force elected officials to make decisions they do not want to make. And no process can prove a guide through tough substantive problems.
GPRA builds on the success of other nations by demanding that government agencies make performance the touchstone for their actions. It provides members of Congress with a way of focusing government on what it does instead of what it spends.
3. Performance Measurement Provides Congress with Critical Information about Agencies’ Strategic Decisions
The fate of the laws Congress passes depends on how agency officials manage them. Members of Congress rightly expect that managers will administer the laws as passed. From the administrators’ point of view, however, the situation is often far more complex and full of conflict. They face competing demands on how aggressively to administer these laws, which can lead to inconsistent decisions, and different laws can ask them to do conflicting things. They sometimes do not have enough money to do all that a program’s beneficiaries might expect, which can produce frequent complaints. And they rarely have enough time to satisfy all legislative goals simultaneously, which means they must make hard choices.
Executive branch officials must constantly make strategic choices about where to invest their energy, how to balance conflicting demands, and how best to achieve the goals they seek. Sometimes they do it like explorers in a jungle, hacking their way with dull machetes through the policy forest. Sometimes they do it much more carefully, planning their steps to maximize legislative goals. One way or another, these decisions determine what gets done and what public programs produce.
Recent rounds of budget cutting have made this process much harder. Downsizing of the federal government’s workforce have hit some agencies harder than others, so those remaining have frequently had to learn how to do new jobs with less help. If budgets have been cut, citizens’ expectations rarely have shrunk, so government workers have faced even harder problems.
This leads to two implications. First, government managers can solve these problems only by thinking and acting strategically. They need to decide how to balance competing demands, which problems demand the most immediate attention, how to apply new technology to do their jobs smarter, and how to deliver more for less. In the midst of such turmoil, government managers often feel they have no time to plan strategically. But in the face of such hard problems, the only possible solution is for them to act and think carefully. The tougher the problems, the more important strategic planning becomes.
Second, congressional authorization and appropriation will increasingly need to build on these strategic plans. The more important strategic planning becomes, the more important it will be for members of Congress to review these plans. They will frame the important tradeoffs that managers deep inside agencies will make. GPRA requires agencies to write strategic plans (which is important in itself), to publish those strategic plans (which is critical to Congress), and to use these strategic plans as the foundation for measuring results (which in the end is most important). This process might seem obscure and esoteric. The transparency it creates, however, can vastly improve congressional oversight, authorization, and appropriations processes because it opens a window into the most important decisions of executive agencies.
The law and logic of GPRA requires government agencies to write strategic plans. These plans will guide members of Congress to the truly critical decisions that these agencies make in implementing the law.
4. Strategic Plans Provide a Road Map to Achieving Results
It is one thing to focus on results. It is another thing to determine how to achieve them. Strategic plans not only provide transparency to the process by which managers set their priorities. They also provide a road map for how managers propose to reach results.
Businesses, of course, have used strategic plans for years, but they are far newer to government. Good strategic plans have six steps:
The Internal Revenue Service, for example, has already developed a strategic plan that charts the agency’s path through these steps. The agency has defined a broad mission:
Mission: The purpose of the Internal Revenue Service is to collect the proper amount of tax revenue at the least cost; service the public by continually improving the quality of our products and services; and perform in a manner warranting the highest degree of public confidence in our integrity, efficiency, and fairness.
It has set three goals:
1. Increase voluntary compliance
2. Maximize customer satisfaction.
3. Achieve quality-driven productivity through systems improvement and employee development
Under the first goal, IRS has specified two objectives:
1a. Collect at least 90 percent of the total tax dollars due and owing, through increased voluntary compliance and enforcement.
1b. Achieve the recognition of the public, outside stakeholders, and IRS employees for the ethical conduct of IRS regarding: fair and uniform application of tax laws; maintenance of the highest standards of integrity; and confidentiality of tax information
The measurable indicators for the first objective is the amount that IRS collects of taxes due and owing. It has set performance goals of 86.6 percent in fiscal year 1995, 86.8 percent in fiscal year 1996, 87.2 percent in fiscal year 1997, increasing to 90 percent by 2001. IRS has also assigned specific responsibility to individual units for achieving these objectives. It has gone through a similar process for the other goals and objectives.
The IRS strategic plan tells everyone:
what mission IRS seeks to accomplish. how IRS plans to meet its mission. what success IRS has had in achieving its mission. who has been responsible for successes and failures.
More than any other tool at Congress’s disposal, strategic planning provides a road map to what agencies are trying to do and how well they are doing it.
More than any other tool at Congress’s disposal, the strategic planning that GPRA requires of government agencies will provide an important road map to what agencies are trying to do and how well they are doing it.
5. Performance Measurement Connects Plans with Results
What agencies do, in the end, matters much less than whether what they do solves the problems for which they were created. It is one thing to forecast the weather; it is another to provide timely information that minimizes the loss of life in a hurricane. The Customs Service can inspect passports and luggage, but that doesn’t necessarily reduce drug smuggling. When we worry about performance, what we really want is to solve problems, not count how much activity agencies have.
This puzzle moves us directly into the arcane distinction between outputs and outcomes. Such talk immediately produces the MEGO (my eyes glaze over) phenomenon. But this distinction matters critically to Congress’s ability to solve the problems for which it legislates, for this reason: Most federal agencies and most federal managers do not directly produce goods and services. Important exceptions include agencies like the State Department, the Social Security Administration, the Department of Veterans Affairs, and the National Weather Service. But most managers spend their time in partnership with other managers—in other federal agencies, at other levels of government, in nonprofit organizations, and in the private sector. It is this partnership that in the end actually produces the goods and services that the government pays for.
For example, the Department of Education actually provides little education; it makes grants to others who educate. The Department of Defense does not build missiles or fighters; private contractors do. OSHA does not make workplaces safe; it regulates and inspects workplaces to ensure that private companies keep their facilities safe. The Health Care Financing Administration does not provide Medicare services; it funds and manages a vast array of nonprofit and for-profit organizations who do so.
This creates a dilemma for measuring results. What matters most is what results federal programs produce. This leads logically to measuring program results. But most of these results depend on how well the partnerships work and therefore are out of the direct control of federal managers. Federal managers understandably are nervous about being measured and held accountable for results they cannot directly control. Instead, they argue that performance measurement ought to focus on their activity. Should GPRA focus on holding managers responsible for results they can control but risk losing sight of the bigger picture, or should it measure the broader impact of government program but risk losing a clear sense of who is responsible for what?
In the language of performance measurement, the question resolves itself to this: Should government measure outputs, the services produced? Or should government measure outcomes, the results achieved?
On the most basic level, the answer is simple. Government must seek to do both. We need information about the specific activities of government managers and about the broader results they produce. But we also need to be very frank: This is very hard to do. The measurement issues at play rank among the toughest technical challenges in public management. The degree of difficulty, however, should not prevent us from trying as hard as we can to do it right. If the solutions are hard, the questions are critical.
On a deeper level, moreover, the return even from basic information can be so great that we ought not be hindered in developing a system just because we cannot fully implement it immediately. The experiences of other nations demonstrate quite clearly that performance measurement emerges only from a very long—probably decades-long—process. In the meantime, however, getting better information about who is doing what, even if that information is rudimentary, can so substantially improve public debate that we ought not be hindered by the scope of the total task.
Does a federal job training program work? It makes much more sense to gauge how many people are trained, what their characteristics are, what kind of jobs they get, and how long they keep them, than simply to declare a program a success because it spends a certain amount of money. Does the air traffic control system work? Comparing delays with on-time arrivals makes far more sense than simply counting outlays. None of these measures gives a full picture. But basic output data is always better than any input data. Advocates of the crime bill several years ago declared success by putting 100,000 more cops on the street. What really matters, of course, is whether the cops actually got to the street and what they did when the got there.
More experienced nations have developed different approaches to these tough questions. New Zealand officials are quite explicit in arguing that the system should be limited to output measures. That, they say, keeps the system firmly grounded and allows clear analysis of who does what. The British government, likewise, has focused on outputs. In Canada and Australia, however, the government has broadened the focus to assessing outcomes, although output measurement remained the basic building block.
We quite simply will not know what works best in the American system until we get more experience. But any output/outcome-based system will be vastly superior to our current non-system, which too often judges performance based on tax dollars spent. GPRA requires agencies to measure outputs; it asks them to move toward outcome assessment. It is thus the first step toward the more sophisticated performance measurement system the nation needs.
The output/outcome question is not an either/or choice. Output measurement is the building block for everything that follows. Moreover, it yields information far superior to input-based judgments. We ought to move aggressively through better strategic planning to output measures.
GPRA builds an important foundation for better policy making by making managers identify and measure outputs.
6. Performance Measurement Can Improve the Authorization Process
Members of Congress face tough questions in authorizing programs: What is the best way to attack a problem? How much money should we spend on trying to do so? What works? Given impossibly competing demands, where should we spend extra money—and which programs should be cut? There are no easy answers to what inevitably requires tough political judgments. But performance measurement can provide some clues.
First, performance measurement generates real information beyond the usual round of claims made by program advocates. What does an agency do and how well does it do it? The experiences of other nations, as well as of state and local governments who have experimented with this system, is that performance measurement greatly informs political judgments. Hard data, moreover, can help counter strong political pressure.
Second, performance measurement provides a way to allocate scarce tax money. Economists talk about the marginal productivity of an extra dollar: Where can we invest the next dollar so that it produces the greatest impact? Performance measurement can identify which programs work best so that members of Congress can tell where money will be best spent.
Third, performance measurement suggests which policy strategies are likely to be most effective. The more agencies describe what they are trying to do, how they are trying to do it, and what they accomplish, the more members of Congress will know about what works. Should federally funded job training rely on federal grant programs to state governments or vouchers to those who need training? Do private companies provide better payroll services than the government’s own operations? Performance measurement can never provide foolproof answers to hard problems. But it can shine a searchlight into the dark recesses of government programs and provide valuable clues about what works best—and why.
By focusing conversations on outputs instead of inputs, on results instead of dollars, GPRA can help members of Congress make better decisions about which programs to authorize and how much money they ought to receive.
7. Performance Measurement Can Improve the Appropriations Process
The Appropriations Committees face even tougher problems: comparing the value of competing programs, and weighing how much to spend overall with the tax dollars available. The struggle revolves around how to create a picture of the whole, how to talk about critical tradeoffs, and how to move the debate from particularized claims to the big policy issues.
As long as the debate centers on how much to spend—that is, on inputs—it will be impossible to jump the gap to the big questions. The United States has spent a decade devising and debating procedural fixes to these questions. Results have proven modest because it is fundamentally impossible to join the symbol of a balanced budget (supported by everyone in theory) with the detailed decisions needed to reach it (contested by everyone in practice). This is a fundamentally unbridgeable gulf.
It makes far more sense to debate what to spend in the context of what results the federal government produces. First, the search for a balanced budget has produced some leverage in making hard appropriations decisions, but it cannot drive all of the tough decisions. Second, a balanced budget might some day be reached, but members of Congress will then need a new instrument to enforce spending discipline. Third, members of Congress are likely to discover that, despite widespread rhetorical support for the idea of a balanced budget, voters might provide little political payoff unless they are also convinced that a smaller government also works better.
Quite simply, Congress in general and the Appropriations Committees in particular need a far more sophisticated tool to make the decisions that lie ahead. The experience of other nations suggests that focusing on results helps provide that tool. It helps crack the door for more competition, and hence more efficiency, in providing government goods and services. It helps build a base for the badly needed restructuring of government agencies and processes. And most important, it changes the very language of the debate.
Information about results has changed the fundamental dynamics wherever performance measurement has been tried. One Phoenix city official told me that it is hard for politically well-connected contractors to push for expensive contracts if the data show that they are twice as expensive as the alternatives. Collecting the information cannot make hard decisions easy. It can, however, fundamentally alter the nature and language of the debate: about what the real issues are, and what the effects are likely to be. I like to think about performance measurement, therefore, as a system of political communication: how we think and talk about the fundamental issues of government.
GPRA can assist the Appropriations Committees by providing a subtle tool to bridge the gap between deficit reduction goals and the decisions needed to reach them. More important, its performance focus can fundamentally alter the terms and language of the debate.
8. Performance Measurement is No Magic Bullet—But It Helps Congress Do What Has to Be Done
It would be easy to overpromise on what GPRA can do and then to be disappointed by what it actually produces. That, after all, has been the track record of previous reforms, like Planning-Program-Budgeting in the 1960s, Zero-Based Budgeting in the 1970s, and Total Quality Management in the 1980s. Frankness demands that we admit two things about GPRA: It is very hard to do; and even if done well, it cannot possibly solve all of our problems.
Indeed, as Australian officials have frankly admitted to me, they have been at the process for a decade and still don’t have it right. But they have also had more than a decade to abandon the process, and they have not. That is because it gives them leverage over important problems that they can get no other way.
Performance measurement will never prove a panacea. It imposes huge technical problems. Quite frankly, it produces few immediate rewards. Doing it well requires time, patience, and investment in new technology and training for the workers developing it. It might well be, moreover, that voters will take excellence for granted and see anything less as a failure. The immediate political payoffs are likely to be modest.
But performance measurement is an inescapable step toward tackling decisions that cannot be ducked. It is, in a sense, the price of admission to the next set of decisions that will face members of Congress: what to do if the balanced budget campaign runs out of gas, and members of Congress need a new tool for gaining leverage over the budget; or what to do if it succeeds, and tough new questions surface—as they surely will. GPRA might be a low-visibility project, but like a strong foundation under a skyscraper, the government’s work is likely to sink and sag even more without it.
GPRA provides no magic bullet for solving government’s problems. But it does provide the foundation for doing what must be done.
9. Performance Measurement Can Transform the President’s Budget Submission to Congress
Because the congressional budget review builds on the president’s submission, the form of that budget can greatly affect Congress’s ability to make critical decisions. If the president’s budget were to be submitted in a performance-based format, it would greatly increase Congress’s ability to ask the important questions. It would also focus the debate between Congress and the president on the issues that matter most.
Office of Management and Budget Director Alice Rivlin has made GPRA the core of her OMB 2000 restructuring effort. OMB has sought to change the focus of its budget examiners from budget requests (inputs) to performance results (outputs). The budget office will have to complete this transition—and so, too, will executive branch agencies—before a budget submission in performance format would even be possible. It might well take the full seven-year GPRA phase-in period to get to a place where a performance-based presidential budget would be feasible. (All federal agencies are not required to submit annual performance reports until March 2000.) But such a performance-based budget surely is a next-round target at which to aim.
GPRA can provide the foundation for even more sophisticated improvements in federal budgeting. Most notably, it could aid the transition to a performance-based presidential budget submission to Congress. Such a budget could enormously improve the ability of the president and Congress to grapple with the truly fundamental budgetary issues.
10. Performance Measurement Can Vastly Improve Congressional Policy Making
Members of Congress today are wrestling with immensely complex problems. And they need the best tools they can get. GPRA is an invaluable addition to the congressional toolbox. In reviewing the steps it requires, it would be easy to mistake GPRA purely as a set of techniques for executive branch managers.
However, GPRA is at its core a tool of greatest use to members of Congress. The very process GPRA creates shines a light on the issues that most demand congressional attention: the choices and tradeoffs executive branch officials make in implementing the law; which policy strategies and tactics work best, and which ones don’t; who is responsible for which results; and where new resources can best be put to use.
In struggling with these issues, members of Congress often find themselves trying to straddle an unbridgeable chasm: on one side, with admirable goals that win enthusiastic support; and on the other side, with tough decisions on the details of policy. What performance measurement can provide is a bridge between them. It does not necessarily make the trip painless, but it can make it possible.
Effective implementation of GPRA therefore ought not to be a goal in itself. That can only promote mindless bean counting. Members of Congress, instead, ought to pursue GPRA because it provides a tool to do what has to be done: grapple with the truly important and fundamental issues that lie inescapably over the horizon, past the balanced budget debate.
GPRA should not be viewed as an end it itself. It provides, rather, a tool to help members of Congress do what they have to do, and to do it in ways that no other existing tool allows.