The state budget should focus on outcomes not allocations

Since 2001, Gov. Mitt Romney and the Legislature have cut spending by almost $3 billion. While also raising taxes and fees, the state has reduced Medicaid services, eliminated health insurance for some of the poor, decimated public health programs such as the anti-smoking campaign and drug prevention, virtually eliminated funding for outpatient mental health care and alcohol and drug treatment, and made deep cuts in higher education, K-12 schooling, and local aid. Local governments and school districts are reeling from the impact, laying off teachers and police officers and struggling to pass overrides to Proposition 2H.

Today an economic recovery is well underway. Yet the most respected independent forecaster in the state, the Massachusetts Taxpayers Foundation, says the state still faces a gap of more than $1 billion between projected revenues and spending in fiscal 2005. As a result, Romney’s 2005 budget proposes further cuts in Medicaid, public health, mental health, corrections, child care, transit, and dozens of other programs.

Welcome to the permanent fiscal crisis.

Like every other state that does not enjoy significant oil or gas revenues, Massachusetts has a long-term structural deficit, a mismatch between projected revenues and projected spending levels. Economic growth will help, but unless we pass deeper spending cuts or larger tax increases, a similar picture will present itself in 2006, 2007, and probably beyond.


The answer lies in the confluence of four basic factors –what MTF has called a fiscal “perfect storm.” First, the rising cost of health care–mounting at a relentless 10 percent a year for the past 25 years–is eating government alive at every level. Nationally, Medicaid rose from $40 billion and 11 percent of state budgets in 1985 to $230 billion and more than 20 percent in 2002. Here in Massachusetts, total health care spending (including Medicaid, insurance for state employees, uncompensated care, and other programs) shot up from 25 percent of state spending in fiscal 2001 to 33 percent in 2004, according to MTF. Romney has recommended further cuts, but even if these are approved, health care would still gobble up $4 out of every $10 of new revenue in his 2005 budget.

Second, the graying of the population–a primary force behind the increasing cost of health care–is also driving up public pension costs. Massachusetts has exacerbated the problem by cutting contributions to the state pension fund for the past three years. This year, facing a crisis, the governor and legislative leaders have agreed to put $530 million into the pension fund, a 77 percent increase that will consume half of all new revenue.

Third, because Massachusetts has a high debt level, interest payments eat up 16 percent of all new revenue in Romney’s budget.

After feeding those three budget busters, as former state senator Patricia McGovern used to call them, Romney has had to cut the rest of state government by 0.2 percent. Given an inflation rate of about 2 percent, that translates into a service-level cut of 2.2 percent.

But there is a fourth element of the perfect storm: a leaky revenue system. With the worst timing imaginable, the citizens of Massachusetts voted themselves an income tax cut in 2000, just as the economy turned south. And like most states, we still do not tax services, which have grown from 40 percent of consumer spending in 1960 to 60 percent today. Imagine if state governments had chosen to tax only agricultural goods when they created their sales taxes, back in the 1930s, so as to protect the new manufacturing economy. We have been just as shortsighted by sparing the service sector today. In the last big fiscal crisis, the Legislature passed a bill to extend the sales tax to services, but Gov. William Weld succeeded in getting it repealed in 1991.

We must maximize the value of every tax dollar.

Since the fiscal storm broke, leaders have grabbed for any solution that could keep them afloat for another year. “Politicians,” says former Minnesota congressman Bill Frenzel, “have more tricks than the CFO of Enron.” Our leaders have been no exception. They have repeatedly balanced the budget using one-time revenues and spending cuts, leaving the same holes to be closed again the following year. According to MTF, the 2004 budget was balanced using $900 million in one-time revenues–which explains most of the structural deficit in 2005. The governor’s 2005 proposal relies on $500 million in one-time revenues and savings, such as those that would come from merging the Massachusetts Turnpike Authority with MassHighway. Even if all these one-timers materialize, we will start 2006 at least half a billion dollars in the hole.

In contrast to the fiscal crisis of the early 1990s, today there is little hope of restoring these cuts through economic growth. At current tax levels health care, pensions, and interest will consume virtually all the revenue growth we can foresee.

To cope with this painful reality, we need reforms that go far deeper than those Romney has proposed. The solutions lie in three places: plugging the leaks in our revenue system, reinventing the way state services are delivered, and changing the way we spend the $25 billion a year we continue to appropriate. In this article, we will focus on the third strategy: reinventing the budget process.

Nothing is more powerful in shaping state government than the way we divvy up our dollars–a process that today drives us inevitably toward across-the-board cuts in spending affecting the poorest and least powerful among us. If we want different answers to our fiscal dilemmas, we must begin by asking different questions.

Most of the Sturm und Drang of budgeting involves the battle over what to cut. Instead, we must learn to focus on what we keep. We should shift the debate away from how best to trim 5, 10, or 15 percent of the budget towardhow best to spend the 85, 90, or 95 percent that remains. We must rethink the way we spend the public’s money, not just to minimize the damage of funding shortfalls but to maximize the value of every tax dollar we collect.

Rules of the budget game

Consider how the budget process works today. The game begins when the budget office receives proposals from department heads. Virtually every budget request starts by proposing to spend what was spent last year. In state budgeting circles, this is often called the “base” or the “current services” level. It represents what is minimally necessary for agencies and departments to keep doing what they’ve been doing the way they’ve been doing it.

But last year’s spending is only the starting point in defining each department’s base budget. Invariably, agencies ask for more money to cover current service levels, on the grounds that inflation makes everything cost more; they may also project higher caseloads or plead that new laws mandate additional services.

Then there are the pre-emptive strikes that departments make to protect themselves from the inevitable cuts. Smart managers build in padding to ensure that, after trimming, they end up with what they feel they really need to operate effectively. Budget staff and elected officials play hide-and-seek with the departments, searching for the padding that can be cut. Meanwhile, policy-makers are bombarded by demands from vocal, entrenched defenders of programs. Unable to tell fat from fundamentals, they often resort to across-the-board cuts, cuts that penalize any administrator who files an honest budget.

What’s worse, this whole game of budgetary cat-and-mouse is played out over programs rather than results. The budget is organized by who spends the money, not what outcomes that money buys. It is policy-making by organizational chart. At its best, this process burns up hundreds of thousands of nonproductive hours. Few people win, many lose, and there is plenty of blame to go around.

Now consider an alternative, which we call Budgeting for Outcomes. It has no concept of “base budget.” Last year’s appropriation is not an entitlement, so the argument is not about adding to or subtracting from it. Instead, budget talks start where the old process ends, with an agreed-upon spending level. Then the discussion revolves around how to buy the best possible results with the people’s money.

Lessons from Washington state

How is this different than the standard budget process described above? Consider the experience of Washington state, the first to use Budgeting for Outcomes.

In its fiscal 2002-03 biennial budget, Washington hit the same fiscal wall that Massachusetts did. Halfway through the biennium, Democratic Gov. Gary Locke and the Washington Legislature were forced to trim $1.5 billion and eliminate 1,340 jobs. But it wasn’t the magnitude of the challenge that frustrated the governor and his staff. It was the futility of the exercise.


“Every step we took, we asked ourselves, why aren’t we asking the right questions? Why are we so focused on the cuts and not on the keeps?” says Marty Brown, director of the Office of Financial Management. “We were missing something. We knew it in our guts.”

Facing another $2.1 billion deficit in the general fund in the upcoming biennium, plus an additional $600 million shortfall in health-services funding, Locke was looking for an alternative to a 15 percent across-the-board cut. He wanted to focus on the big questions: What should state government do–and what should it stop doing?

“I don’t want to thin the soup,” Locke declared at the time. “I want state government to do a great job in fulfilling its highest priorities.”

‘Why are we so focused on the cuts and not the keeps?’

In August 2002, Locke’s chief of staff asked our company, the Public Strategies Group, for help. We made him a proposal that was anything but reasonable. PSG proposed to start with the results citizens wanted, not the programs the agencies operated. We urged the governor and his staff to focus not on how to cut 10 to 15 percent but on how to maximize the results produced with the 85 to 90 percent remaining. Locke decided that this unreasonable approach was the only reasonable thing to do.

PSG helped the governor’s budget staff design a process to answer five key questions:

  • Is the real problem short-term or long-term?
  • How much are citizens willing to spend?
  • What results do citizens want for their money?
  • How much will the state spend to produce each of these results?
  • How best can that money be spent to achieve each of the core results?

These five questions led to five key challenges.

  1. Get a grip on the problem.

    How you define a problem dictates how you approach the solution. Washington’s fiscal staff defined the problem as the convergence of three forces: a deep economic recession that slashed revenues; permanent limits on revenue and spending growth imposed by anti-tax activists through statewide initiatives; and rising costs for the core activities of the state (“education, medication, and incarceration,” as Marty Brown describes them). Of the three, only the recession’s impact on revenue could be termed cyclical, likely to turn around at some point. The other two were more or less permanent. Thus, the solutions had to be more or less permanent.

  2. Set the price of government.

    This was the purview of a Guidance Team, made up of senior policy people, plus several leaders from business and private think tanks. (Organized labor was invited to participate, but chose not to.) Its first big decision was to build the budget based on expected revenues under existing law, without new taxes. In November 2002, despite heavy lobbying by Locke, voters soundly defeated a gas tax increase to pay for long-needed transportation projects. This anti-tax reality–plus a fear that tax increases would further depress the state’s economy–led the team to advise the governor against raising taxes.

  3. Set the priorities of government.

    Working with a team of senior staff from the state’s Office of Financial Management, the Guidance Team defined the results it believed Washington’s citizens most wanted from state government, boiled down to 10 “Priorities of Government,” as the governor declared them. These top priorities called for improvement in:

  4. student achievement in elementary, middle, and high schools;
  5. the quality and productivity of the workforce;
  6. the value of a state college or university education;
  7. the health of Washington’s citizens;
  8. the security of Washington’s vulnerable children and adults;
  9. the vitality of businesses and individuals;
  10. the statewide mobility of people, goods, information, and energy;
  11. the safety of people and property;
  12. the quality of Washington’s priceless natural resources; and
  13. cultural and recreational opportunities.
  14. Allocate available resources across the priorities.

    The next challenge was allocating the state’s entire $54 billion budget ($24 billion in the general fund and $30 billion in other funds) across the 10 priorities. Setting aside 10 percent for overhead functions, such as pension contributions and administrative services, the Guidance and staff teams parceled out the rest among the 10 desired results, trying to use a citizen’s point of view based on perceived value, rather than on past practice. In some areas their choices reflected existing patterns of spending, but in a few they made changes–allocating more resources to economic vitality, for example, and fewer to public safety.

    This wasn’t easy. The teams wanted data on past spending patterns. We told them there wasn’t any, because budgets had always been organized by agency costs, not by the value of outcomes. On top of that, multiple programs and agencies contributed to each result. We asked them to put a price on each outcome, to express the relative value of each one.

  15. Develop a purchasing plan for each desired result.

    These state leaders then put together Results Teams–one for each priority outcome, made up of knowledgeable people from agencies involved in that policy area. “We asked them to forget the loyalties they have to the agencies they represent,” said Locke. “‘Be like citizens,’ we said. ‘Tell us where to put the money, so we get the best results. Tell us what similar programs can be consolidated. Tell us what programs don’t make a large enough difference in getting the results we want.'”

    The teams each chose three indicators that would measure progress toward their goals. Then they developed strategy maps, or explicit cause-and-effect diagrams showing what they considered to be the best ways to achieve the desired outcomes. And they had to articulate their “theory of what matters most”–how much the different activities they favored would contribute to the desired outcome.

Since the 10 priorities of government cut across departmental lines, so did the Results Teams. That led to predictable challenges. The mobility team, for instance, had responsibility for improving the movement of people, goods, information, and energy. But many of its members, with backgrounds in highways and public works, were accustomed to focusing on roads and bridges. As they wrestled with their charge, however, they came to see that the challenge was as much about telecommuting by World Wide Web as commuting by car. Ultimately, the team came to understand the fundamental distinction between pouring cement and moving people, ideas, goods, and energy efficiently.

Creating a strategy map required those involved to take a firm position on how activity adds up to results. Armed with a “theory of what matters most,” each Results Team identified five or six key strategies for producing its desired outcomes. This process stimulated a kind of creativity that is absent from traditional budget development. For example, the team dealing with K-12 education said the state needed to purchase more early childhood education, shift to a “pay for skills” compensation system for teachers, and move away from across-the-board school funding toward targeted funding for those schools and kids most in need.

There is no one right way to create a cause-and-effect map and translate that diagram into funding priorities. But here is how the health team did it, beginning with their map (see below).

To achieve improved health for the citizens of Washington, the health team identified four possible strategies: increasing healthy behaviors among citizens (eating better, driving more safely, quitting smoking, and getting more exercise); mitigating environmental hazards (making water, air, and food cleaner and safer); identifying and mitigating risk factors related to gender, socioeconomic hardships, and genetic predispositions; and providing access to appropriate, high-quality medical and mental health treatment.

In ranking these strategies, mitigating environmental hazards was declared the most important, followed in order by promoting healthy behaviors, providing access to health care, and mitigating risk factors. Given the state’s fiscal crunch, the team decided to give funding priority to the first two, even though that meant reduced spending on traditional (and expensive) medical care. In fact, the team’s analysis of available data showed that these two strategies would yield a 16-to-1 return on the state’s investment.

The old budget game would have given priority to the area of greatest spending–medical treatment–by protecting it from cuts as much as possible, even if that meant shortchanging health promotion. (That’s exactly what happened in Massachusetts, as the state all but eliminated prevention programs, while Medicaid spending continues to rise, even with service cutbacks.) But the new approach asked the health team to ignore last year’s spending and figure out where the best results could be obtained for the available money. The “aha” moment came when a PSG consultant asked why team members weren’t emphasizing personal choice and behavior by focusing on drugs, alcohol, and obesity. “Oh, we could never say that out loud,” the team leader responded. But the consultant pushed back: If they truly thought investment in prevention would produce more health improvement than a comparable investment in treatment, they should say so. They did–and elevated health promotion over medical treatment for the first time in the state’s history.

But the budget wasn’t done yet. Next, the Results Team leaders met to talk about what they could purchase from one another to achieve their goals. The higher education team decided to use some of its funds to pay for better K-12 education, to better prepare its incoming students. Two teams jointly bought increased efforts to protect water quality, to improve both health and natural resource outcomes. Because the Guidance Team had allocated resources to improving safety that fell well below the current costs for incarceration, the safety team planned to focus resources on the most threatening criminals and release 6,000 low-risk felons early. The economic vitality team felt that releasing so many offenders would adversely affect the state’s ability to attract business, however, so it contributed money to keep more of them in jail. This cross-team buying was necessary because the work of state government is interconnected: Spending in one area contributes to outcomes in others.

Finally, the process turned to existing state activities–the place a traditional budget process starts. Each Results Team was given a subset of the 1,300 state activities funded by the traditional budget. “Their mission,” the governor explained, “was to get more yield on less acreage.” To do so, they put together a detailed purchasing plan, indicating:

  • what they would buy–both new and existing activities;
  • what else they would buy if they had more money;
  • what they would eliminate first if they had less money; and
  • what they would not buy.

Using these and similar rankings provided by the agencies, the Guidance Team and other advisors made final recommendations to the governor–giving him, in effect, 10 strategic programs for state government that linked results, indicators, strategies, and purchase plans.

The governor generally followed these purchase plans in his budget proposal. Under each of his 10 priorities for government, his budget showed those activities that would be funded and those that would not–as the graphic at right demonstrates. It was clear and easy to understand, and it explained in simple terms why some activities continued and others were eliminated.

No pain, no gain

That’s not to say that the fiscal 2004-05 budget, crafted at a time of extreme fiscal hardship, came easy. Locke had warned that his budget plan would be painful, and it was. Locke proposed to eliminate health insurance for nearly 60,000 of the working poor; dental, hearing, and optometric coverage for poor adults on Medicaid; and 2,500 state jobs. His budget eliminated cost-of-living increases for state employees and suspended teacher pay increases and a $221 million class-size-reduction effort, both mandated by citizen initiatives. It also called for university tuition to rise by 9 percent a year for two years; for 1,200 low-risk felons to leave prison early; and for a number of smaller programs to be shut down altogether.

But rather than being denounced as a slash-and-burn exercise, Locke’s budget was accepted with relative equanimity–and, more important, followed. As Joe Dear, the governor’s former chief of staff, put it, “Never has such bad news been received so well.”

“Gov. Gary Locke’s budget is a big step forward for Washington,” declared the Seattle Times. “Few Washingtonians will find much to like about the brutal state spending plan Gov. Gary Locke recommended Tuesday,” added the Tacoma News Tribune. “But as ugly as the result was, there’s a lot to like about the way Locke and his staff arrived at it, using a new process that forced hard choices about the core priorities of state government.”

Locke had not always been treated with this kind of deference. In his 2000 re-election campaign, Locke was slammed by John Carlson, his Republican opponent, for lack of leadership. Soon after Locke’s new budget was released, however, Carlson wrote a column declaring it “a work of bold, impressive statecraft.” Carlson told the Seattle Times: “Genuine leadership is doing what must be done when you don’t want to do it. And I think the governor is doing that.”

Voters agreed. In a late January survey, 64 percent of respondents endorsed the following statement: “Whether or not I agree with all of the governor’s budget recommendations, I respect his leadership and vision to solve the current problem and get the state’s economy back on track.” Only 29 percent disagreed.

The Legislature, which is split between a Republican Senate and a Democratic House, also liked the new budget format. “It was astounding,” says Marty Brown, the state finance director. “I’ve never been to a set of hearings where the reception was so positive, despite the amount of bad news we had to deliver.”

In early April, when the Republican majority in the state Senate presented its own budget, the first slide was titled following the governor’s lead. Despite deep differences between the parties over taxes and budget cuts, the Legislature ultimately passed a budget that was remarkably close to Locke’s proposal. Legislators approved Locke’s early implementation of new sentencing guidelines that allowed 1,200 prisoners to be released, as well as his proposed delay in voter-approved initiatives to reduce class sizes and grant automatic pay increases to teachers. They also agreed to amend another public initiative-backed plan to expand coverage in the state’s basic health plan, so that funding could go toward current programs. And they required that the next biennial budget be structured around the 10 Priorities of Government, with outcome measures to track progress against each one–along with outcome measures for each activity.

Unfinished business

In remaking its state budget, Washington has more to do. Some questions were too big to be tackled in one crisis-wracked budget cycle. Locke has proposed a joint legislative-executive study of the K-12 financing system, for instance, to examine the options more carefully and build the political support necessary for reform.

Nor did the Locke administration have time for one of the final steps in the Outcome Budgeting process: soliciting offers to produce the desired results from all comers, public and private. Washington’s Results Teams treated the past budget’s 1,300 activities as the available universe of offers. But to maximize the impact of Budgeting for Outcomes, purchasing teams should solicit offers to see who can deliver the most results for the money.

This is the step that departs most radically from the old budget game. The purchasing teams incorporate the outcome, indicators, price, and purchasing strategies they have settled upon into one or more solicitations–let’s call them “requests for results.” The RFRs can be sent to all agencies and departments, to other governments, and to nonprofit and for-profit organizations. They ask each of these potential suppliers to identify how they would help deliver the expected results, and at what price.

In developing their responses, bidders need not, indeed must not, take anything for granted. They must assume that for each result there will be many proposals from many potential sellers, public and private. If they expect to get funded, they have to offer up proposals that deliver the needed results at a competitive price. Since one organization may choose to submit multiple proposals (for its various programs and activities), it is in a sense competing against itself. This, too, forces it to challenge its own practices, to examine whether they are the most cost-effective means to the end.

While the process is challenging to bidders, it also liberates them. They need not be limited by the past; the process encourages them to come up with new approaches and creative twists. Some will forge partnerships with other departments or agencies, with other governments, and with nongovernmental organizations. All will have incentives to use the many tools that help squeeze more value out of every dollar, such as rightsizing, consolidation, competition, information technology, customer choice, Total Quality Management, and Business Process Re-engineering.

Once the offers are in, purchasing teams rank them in terms of results delivered per dollar, moving down the list of desirable offers and buying according to priority until available funds have been exhausted. This buying plan becomes the budget proposal. It is a list of keeps, not cuts: positive choices for spending the citizens’ resources to buy the citizens’ results.

Those items ranked too far down the list to be purchased don’t make the budget. For defenders of these programs, the challenge is clear: Improve your ranking. Demonstrate that your program will deliver better results (or the same results at a better price) than others being offered, or demonstrate that there are more-efficient ways for those at the top of the list to deliver their results, thereby making room for more purchases. Here again, the incentives are as they should be: better results for citizens, at a better price.

Also underdeveloped in Washington is the role of the Legislature in Outcome Budgeting. For fiscal ’04-05, Locke developed his own budget proposal and submitted it to the Legislature. Ideally, with more time to prepare, the governor and legislative leaders would agree in advance on the price of government–what percent of personal income state government would raise in taxes, fees, and charges, and thus how much revenue it would have to spend. The two branches would also work together to define the outcomes most important to citizens. With that as a starting point, the governor and his budget staff would develop their budget proposal. Over time, the Legislature could shift its committee structure to reflect the key outcome goals, paralleling the administration’s purchasing teams.

Data on results will never crowd out politics and interest group pressures in an elected legislature, but it can enter the mix. One way to make that happen is to involve legislators in the process of creating the outcome goals, so they develop some sense of ownership. Another is to demonstrate that improving outcomes matters to their constituents. Some states and nations publish scorecards showing annual progress on key goals, for example–scorecards that get significant attention from the media and the public.

Since the state of Washington pioneered Budgeting for Outcomes, the idea has begun to spread. Iowa Gov. Tom Vilsack and Los Angeles Mayor Jim Hahn, both Democrats, have embarked on the process. But in Massachusetts, the budget process for fiscal ’05 is well underway in traditional fashion, with its familiar bureaucratic–and political–attributes. Gov. Romney’s budget is based largely on this year’s spending plan, his priorities expressed only at the margins through incremental expansions and relative cutbacks. His “reforms” are mostly proposals for bureaucratic reshuffling that promise dubious savings. The Democratic Legislature stands poised to reject most of them and to substitute its own grab bag of funding preferences, but they differ from the governor’s only at the edges. Meanwhile, the great beast of the state-government status quo lumbers on, unaltered and unchallenged. Absent from the entire exercise is any serious rethinking of what government does or how it does it.

This should come as no surprise. In government, as in most realms of life, we find it safer and easier to do what we’ve always done, simply because we’ve always done it. Only on rare occasions–usually in times of crisis–do we step back to gain a broader perspective, erase all our preconceived ideas and routine behaviors, and take a fresh look at how to make the most of our limited time and resources. But if we, and our elected leaders, do not take the opportunity of fiscal crisis for this kind of rethinking, when will we do it?

Meet the Author
Meet the Author
Budgeting for Outcomes allows public leaders to do some of this big-picture, creative thinking each time they prepare a budget. In fact, the process demands it. The challenge for departments shifts from padding their base to proving that their programs produce desired outcomes for the best price. The job of the budget office shifts from playing truth-or-dare with department heads to comparing competing strategies to determine which ones offer the most bang for the citizen’s buck. For everyone involved, from elected officials to budget analysts to department heads, being a public servant suddenly focuses on delivering results that citizens value at a price they are willing to pay. Is that not a change worth making?

David Osborne and Peter Hutchinson are co-authors of a new book, The Price of Government: Getting the Results We Need in an Age of Permanent Fiscal Crisis, on which this article is based. Osborne, of Essex, Mass., is a senior partner and Hutchinson is founder and president of the Public Strategies Group (, based in St. Paul, Minn.