Fundamentals of Sound State Budgeting Practices

June 1995

Introduction

This report deals with the nuts and bolts of state budgeting practices--ways to make the process of enacting and managing a budget work more smoothly. It is designed primarily for new legislators, legislators who have not been members of budget committees, and people outside legislatures who are interested in state budgeting processes. The report does not try to explain state budgeting processes from A to Z. It is concerned with some issues in the budget process that are common to all the states and it reports on procedures that states have developed to strengthen the process.

Where it seems appropriate, this report recommends specific practices. The practices recommended here cannot be expected in and of themselves to end revenue shortfalls, ensure balanced budgets, or settle differences over policy. They can be helpful in eliminating procedural issues and allowing policymakers to focus on issues that need attention, such as what the state really should be spending the taxpayers' money on. Budget techniques cannot respond to this kind of question. Although the central purpose of every state's budget process is how to allocate fund, there is no single, preferred procedures. As a result, the main principle of sound state budgeting is to maintain flexibility.

"The power to tax involves the power to destroy," said Chief Justice John Marshall, and it is equally true that the power to spend is the power to create (note 1). Budgets are documents that express state governments' power to act. They summarize policymakers' evaluations of past programs and public agencies and their forecasts of current and future needs and resources. Budgets set goals, decide among alternative objectives, and create means for controlling and accounting for the expenditures of public money. They can create pressures for tax increases or tax cuts. They can push reform or they can discourage it.

Because budgets have so many functions, the process of writing one is often conflict-ridden, unsatisfactory to observers and participants, and flawed in its outcomes. Budgets seem to increase rather than resolve partisan competition; they sometimes are late; they lave problems unresolved; they spend too much or too little; they may fail to include adequate program review, planning for the future, accounting for past expenses, or controls on planned spending (note 2).

These complaints have shown up ever since formal, comprehensive budgeting became a feature of state and local government in the early years of the 20th century. The Taft Commission, which examined federal budget processes in 1912, criticized federal budgeting procedures for the same flaws observers note today. Some of the problems--partisanship, indecisiveness, lack of closure--are inherent in the democratic process. Others spring from conflicting expectations of the process. The central function of a budget--the decision of how much to spend for what--will always create disputes, and no budget will ever satisfy everyone.

This document highlights some central issues in the state budgeting process, summarizes current thinking on them, and identifies some mechanisms and techniques that can help solve problems in the process.

The Challenge to Traditional State Budgeting

This report begins with a discussion of alternative ways of thinking about and writing state budgets. Traditional state budgeting is under attack on the grounds that its processes are fundamentally flawed. Critics charges that current budget practices discourage examining program outcomes in favor of automatically continuing existing programs, discourage policy analysis in favor of legislative micromanagement, and are too tolerant of bureaucracy and waste. They allege that budgets fail in their mission of reviewing the past and planning for the future.

Chapter 1 looks at these issues:

• Traditional methods of state budgeting
• Performance-based budgeting
• Zero-base budgeting
• Why traditional budgeting survives

Budget processes always benefit from reconsideration because, at the very least, reconsideration improves people's understanding of the difficulty of the process. This chapter contends that changes in budgeting processes will not achieve the desired goals if too much is expected of them. That is not an argument against change. It is a caution against jumping on bandwagons.

Traditional methods of state budgeting

Participants and observers agree that state budgeting is less satisfactory, in process and in outcome, than in previous years. Hard times always make for hard budgets, but many feel that in recent years, the process itself has been to blame, particularly because of its traditional focus on line-item control and incremental budgeting.

Traditionally, state budgets have focused on controlling expenditures. Control is expressed in written budgets through "line items"--statements allocating so much money for a specific expense: computers for the tax collectors, acquisitions for the state library, salaries for prison guards. Where written budgets focus on line items, legislators tend to do so as well.

Line-item budgeting tends to be incremental--previous appropriations are increased or decreased by small increments over time. This approach is likely to take previous policies and programs for granted, and discourages rigorous or fundamental review of priorities, program effectiveness, or service outcomes.

These practices are under attack because they are said to foster a business-as-usual approach to government at a time when the public is challenging how state governments operate, questioning their efficiency and effectiveness, and expressing distrust of representative government itself.

Line items focus on what money buys (an input) rather than on the service that is provided (an outcome). An input could be tones of asphalt for state highways or new computers for property tax assessors. Outcomes are less likely to be considered:

Have traffic conditions improved enough to justify the asphalt? Are property tax assessments more timely and equitable? Nothing in a line-item budget prevents such questions from being asked, but the format does not encourage questions because the written form of budgets tends to drive the way people think about them.

With growing concern about how well government functions, many people contend that the traditional focus on line-item budget and incremental change neglects outcomes so much that the budgeting process itself is an impediment to effectively delivering programs. Critics contend that line-item budgeting does not do enough to take program results into account. Two questions arise: Are these criticisms justified? Can adopting of a different technique improve budget outcomes? This chapter first looks at two proposals for reforming budget processes--performance-based budgeting and zero-base budgeting. Then it returns to the evaluation of traditional incremental budgeting.

Performance-based budgeting

Governors and legislators throughout the states are revising budget procedures to emphasize performance and results. This activity goes under many names: outcome-based budgeting, performance budgeting, and sometimes program budgeting. The terms are confusing because in current use they overlap but do not mean exactly the same things. In general, though, the present trend is to reshape budget processes to reward efficient, effective programs and to encourage remodeling programs that cannot meet specific goals. In this report, this kind of budgeting is called performance-based budgeting.

Performance-based budgeting calls for a revolution in how states are governed. It focuses on setting goals, designing the strategies needed to meet the goals, and measuring how well they are met. Future funding decisions should focus on program effectiveness, not on the preservation of existing programs and levels of spending. This approach requires that budgeting be directed at program rather than at specific line items, that the goals of those programs be laid out in measurable terms, and that performance review becomes central to budget decisions (note 3).

None of this is easy:

It is hard to identify and reach agreement on quantifiable goals for most state programs. For example, in the ambitious performance-based budgeting that Governor William F. Weld of Massachusetts developed for fiscal year 1994, the agency and program mission statements explain what the agencies do rather than set goals.

Any reasonable definition of state programs has to recognize that many of them cross existing agency lines, so that full implementation of a program-oriented performance budget might require extensive reorganization. The Legislative Analyst in California has in fact suggested that effective program management will require restructuring the division of responsibilities between state and local government, which would require constitutional changes.

Advocates contend that the difficulty of implementing a performance-based budget is evidence of how thoroughly state government needs to be reformed. They say that the difficulty of agreeing on goals for programs is evidence that the issue has been neglected, and the process of trying to reach agreement will produce valuable analysis and debate. If existing agency structures impede a focus on programs and service delivery, the framework needs rebuilding. The difficulty of measuring performance has to be faced squarely. How else can anyone know whether government is providing needed services? How else can public confidence in government be rebuilt?

Massachusetts Governor Weld submitted his budget proposals for fiscal year in the form of a performance-based budget, in many ways providing a working model of the general purposes of such reform. His budget proposals and the appropriations bill he submitted to the legislature are organized by agency programs, without line items below the program level.

All programs--service delivery groups in Governor Weld's idiom--developed performance measures including such categories as the amount of service provided, productivity, degree of public satisfaction, equity of service, delivery, and the extent to which a program met its immediate purpose. The budget followed up this logic by proposing greater authority for agency managers to transfer funds from one program to another under their control. Massachusetts legislators were unwilling to adopt much of Weld's restructuring of the budget, in part because it would have increased executive authorship at the expense of the legislature.

Texas legislators have adopted a performance-based budget, and in Texas's case, the Legislature so thoroughly dominates the budget process that there is little likelihood of any increase in executive control. The budget for the current biennium required six-year strategic plans from state agencies and provided funding according to the agencies' goals and objectives.

There is widespread enthusiasm for performance budgeting in the states, and some states have implemented some elements, especially the goal-setting and the greater attention to performance measurement. Questions remain:

• What will a state legislature do if a program does not reach its target goals?
• How can rewards for performance be established without creating incentives to reshape programs to reap rewards rather than improve programs?
• How willing are legislators to trade certain control over budget detail for promises of improved service delivery that require greater executive discretionary power?

Zero-base budgeting

The popularity of zero-base budgeting (ZBB) is partly due to its name. It appeals to many people who are concerned with public budgeting because, according to one standard definition, it requires "the review of all budget requests from point zero, without assuming that any existing program should continue" (note 4). Although the original goals of ZBB have proved elusive, in a modified form it has become a widely used budgeting tool.

In its original sense, ZBB meant that no past decisions are taken for granted. Every previous budget decision is up for review. Existing and proposed programs are on an equal footing, and the traditional state practice of altering almost all existing budget lines by small amounts every year or two would be swept away. No state government has ever found this feasible. Even Georgia, where Governor Jimmy Carter introduced ZBB to state budgeting in 1971, employed a much modified form.

State programs are not, in practice, amenable to such a radical annual re-examination. Statutes, obligations to local governments, requirements of the federal government, and other past decisions have many times created state funding commitments that are almost impossible to change very much in the short run. Education funding levels are determined in many states partly by state and federal judicial decisions and state constitutional provisions, as well as by statutes. Federal mandates require that state Medicaid funding meet a specific minimum level if Medicaid is to exist at all in a state. Federal law affects environmental program spending, and both state and federal courts help determine state spending on prisons. Much state spending, therefore, cannot usefully be subjected to the kind of fundamental re-examination that ZBB in its original form envisions.

As practiced in state government, ZBB has been used to consider reducing activities in a program, not just to look at its continuation or expansion. It does so by requiring that

• agencies' formal budget requests include the consequences of different funding levels, and that
• activities be prioritized according to the possibility of increased and decreased funding for each of them.

ZBB appeared in the 1970s when rapid growth in federal programs, grants-in-aid to states, and state tax sources seemed to be bringing inexorable and unexamined growth to state budgets. Hence it focused on potential restraint and possible cuts, rather than on the quality of service delivery. The first two states to adopt ZBB enthusiastically were Georgia and Texas, both of which were at the time experiencing economic booms and above-average expenditure growth rates.

In the states where it continues to exist in some form, ZBB is hard to separate from the widespread practice of expecting agencies to evaluate the impact of changes in funding on operations. This is a useful technique. It provides valuable information both when state resources are expanding and when cuts are needed, and assists policymakers to break with the tradition of incrementalism. Even more important in the budget climate of recent years, the process makes it possible to avoid across-the-board cuts by emphasizing the effects of different cuts on services.

To the extent that ZBB has encouraged governors and legislators to take a hard look at the impact of incremental changes in state spending, it produced a significant improvement in state budgeting. But in its classic form--begin all budget evaluations from zero--ZBB is as unworkable as it ever was (note 5).

Why traditional budgeting survives

Why is it so hard to change traditional budgeting methods? Why does budget reform seem to excite more interest than activity? Critics of proposed reforms suggest two reasons:

• Traditional budgeting meets more expectations about the process better than any proposed reform, and
• Reforms will not solve the problems that reformers have identified.

Proposals for reform focus on particular unsatisfactory results from the existing process and recommend ways to improve those results. But they may fail to consider how many conflicting expectations the budget process has to meet. Aaron Wildavsky, an expect on government budgeting methods, puts it like this:

Budgeting is supposed to contribute to continuity (for planning), to change (for policy evaluation), to flexibility (for the economy), to rigidity (for limiting spending) and to openness (for accountability) (note 6).

Wildavsky goes on to suggest that a budget process that is expected to do so many disparate things will work worse as more specific formulas and expectations are loaded onto it. Budgeting, he argues, should not be made responsible for all of the aims of government. Traditional budgeting has responded fairly well to the conflicting demands made upon it because it builds upon previous agreements and commitments. It does not reopen every question and it does not try to do too much. Planning, evaluating, and accounting are activities that can proceed effectively without being central to the budget. For budget processes not to be overloaded, they should continue to focus on narrow, not broad, purposes.

Hal Hovey, editor of State Policy Reports and a close observer of state reform, concurs, writing recently in The Fiscal Letter:

Many of the values of reforms can be lost by expecting too much from them. They won't ever solve the real problem, which is that we voters want to spend more than we want to pay in taxes, and insist on elected officials who agree with us. We are all in for trouble if state officials do what the Congress has made a practice of doing--substituting a new round of budget reforms for dealing with the budget (note 7).

Both Wildavsky and Hovey are suggesting that traditional budgeting survives because of its lack of rigorous method. The traditional systems allows but does not require legislators and governors to use many techniques and kinds of information now available and similar to those which reformers urge them to adopt. Program and financial auditing, performance evaluation teams, and short- and long-term forecasting already are institutional fixtures of many state governments. Sunset review processes also exist in many states. There is no lack of review. What is lacking, according to legislators and the staff who provide the audits and reviews, is the use of the information in budgeting.

Much good information goes unused because legislators sometimes find program reviews and audits focused on the wrong question, too detailed, or untimely. Some issues can be addressed with better communication between legislators and evaluators. A more significant reason is time itself: Legislators lack the time to make use of all the information that is available to them. The lack of rigorous method in traditional budget processes allows legislators to focus on their priorities to the exclusion of others, work on what is most important at the time, and rely on the continuation of past decisions in other areas. This flexibility is the greatest strength of traditional budgeting methods.

Constraints on State Government

At the same that that management theorists are recommending bottom-up management accompanied by maximum flexibility at the top, many state governments face new constraints on their flexibility. Tax and expenditure limits have become widespread, and a growing number of states require voter approval of tax increases. Policy decisions are increasingly subject to voter and judicial review. Federal officials continue to enact more mandates, and taxpayers are unwilling to pay for more government.

Budget processes cannot control or even limit voters' distrust of elected officials, federal policies toward states, or the outcome of lawsuits challenging state corrections, education, or health policies. Budget processes have proved unable to control rapid growth in expenditures for corrections, education, and Medicaid. In many states, major political issues are decided outside of legislative bodies--by courts, Congress, economic forces, and voter initiatives. In this environment, it is sensible for legislators and governors to maintain as much flexibility as possible in their area of decision making.

This chapter describes the most important kinds of limits recently imposed on the flexibility of state government, and discusses some ways to preserve the flexibility that remains.

Tax and expenditure limitations

By mid-1994, 21 states had adopted legal caps on the growth of state expenditures or revenues to control budget growth. This "first generation" of state expenditure limits generally failed to control the growth of state budgets. The "second generation" of expenditure, which require voter approval of tax increases, seems more likely to stabilize or even reduce state government expenditure. But so far, there seems to have been no significant difference in the amount of budget growth between states that have and states that lack expenditure limits (note 8). Where limits on local governments expenditures exist, they may in fact have encouraged the growth of state budgets by shifting responsibilities for administration, funding, or both from local government to state government.

Expenditure limits themselves have been limited in their effects by loopholes for important expenditures like capital funds, special-purpose funds, and federal grants. In some cases, spending caps have been set at a level so high that they have never come into effect. A 1989 study of 12 states' expenditure caps found that nine of them had never been triggered (note 9).

Spending limits have, however, influence how legislators and governors think about budgets. Some observers contend that the existence of limits induces caution within state government and constrains budget requests and appropriations. Overall, though, spending caps are likely to be less effective than the revenue limits that voters have begun to impose, such as referral of state tax increases to the voters.

It is too early to measure the impact of the rigid limits on taxes and expenditures that Oklahoma and Colorado voters added to their state constitutions in 1992. Colorado's "Amendment 1" requires voter approval of all state and local government tax increases, making cost shifting between state and local government virtually impossible. Oklahoma's amendment effectively requires voter approval of state tax increases, and in the opinion of some observers will slowly cause local governments and users of state services to pick up more costs.

A more certain impact will come from the current trend to substitute state tax collections for local property tax support for schools--a decision that even in a relatively small state like Oregon can require state spending to grow by hundreds of millions of dollars a year. Michigan has enacted $3.4 billion in new state taxes to fund such a shift. Not all voters are as willing as those in Michigan to accept state tax increase in proportion to the cut in local property taxes, as Oregon voters have demonstrated in turning down state tax increases to replace property tax cuts. Such voter mandates are likely to mean substantial pressures on other areas of state spending, and substantially reduced flexibility for those who write state budgets.

Maintaining flexibility within tax and expenditure limitations

The following suggestions are appropriate for any legislator to consider, since their purpose is to sustain the flexibility that is an essential component of an adaptable and responsive legislative budget process. They become all the more important when tax and budget limitations eliminate some other possibilities.

1. Review the division of responsibilities between state and local governments.

The old principle that responsibility for funding a program goes along with power to operate it has become less and less evident in state and local government, making it harder for taxpayers to know which government does what and who pays for what. One way of clarifying the connections between taxes and services is to sort out the responsibilities of different levels of government (note 10). It may be possible to return locally administered programs to local funding. The current trend to reduce local property taxes for schools appears to be a school finance issue, not a general opposition to local taxes. Voters continue to have far more confidence in local governments than in state governments (note 11). Colorado's experience to date with the constitutional requirement for voter approval of all tax and fee increases is that voters largely support well-reasoned proposals from their local governments, with voter disapproval more common as the government--special district, school district, municipality, county, or state--seems larger and more distant.

2. Consider imposing program costs upon program beneficiaries.

Local governments increasingly rely on fees and charges as a source of revenue. California municipalities in particular have done so since the property tax limit was approved by voters in 1978; 28 percent of the revenue local governments raised in California came from charges in 2990, up from 15 percent in 1978. State governments also have increased their reliance on fees and charges, but to a smaller extent. In part this is because charges are less workable for state services than for municipal and county services. Education, recreation, transportation, environmental preservation, and licensing and regulatory activities may be appropriate areas for cost transfers to users.

There are possible disadvantages that have to be considered case by case. Imposing costs on users discourages the use of a service--higher fees for college classes can cut enrollment back. Discouraging use of a service may or may not be desirable: People who advocate charging drivers for using highways during peak commuting hours contend that the reduced congestion would provide drivers with savings. A different consideration is that agencies or programs that charge fees for their services often come to feel that the money raised is their own and that they should be allowed to decide how it is spent. It may take a special effort to continue legislative oversight of the use of revenue from fees and charges.

3. Avoid earmarking revenues.

Earmarking revenues for designated expenditures merits careful examination. The argument for doing so is to provide continuous, guaranteed support for a favored program, or to win voter support for a tax increase. The argument against earmarking is the reduction in decision makers' power to set budget priorities. A revenue that is earmarked for a particular expenditure may provide too much or too little revenue; a good fit is hard to make. If it is too little, there is no reason for the earmark ring; if it is too much, it can be impossible to shift resources to areas of the budget with higher priorities. The long-term consequences of earmarking is reduced control over the relative shares of state spending for different programs, a kind of rigidity that is exactly opposite to the flexibility and program review many encourage state governments to increase. Similarly, mandatory expenditure requirements (a requirement that a given percentage of total revenue growth or collections go to a specified program) hampers good budget management and adaptability.

These general rules have to be modified to fit circumstances. State governments have followed the examples of local governments in relying more on fees and charges for services to pay for the services, a practice that virtually requires earmarking if users are to be assured that university fees fund university services, tolls pay for specific highways, and admission fees help to improve parks. More significant in state finance overall has been the reappearance of explicitly or implicitly earmarking sales or income tax increases for elementary and secondary education, as enacted in Arkansas and Oklahoma in recent years and as proposed in Colorado in 1992. Such decisions are a tradeoff between flexibility in budget management and maneuvering to enact a tax increase.

4. Review fund management practices; consolidate funds in the state general fund.

Having a large number of separate funds is a relic of 19th-century state budgeting, when the practice was to assign a revenue source and a fund to each of many different activities and to get along without a comprehensive budget. A large number of funds unnecessarily complicates revenue forecasting, budgeting, and accounting, and is likely to confuse the public. Money management can be a profitable revenue source for state governments; consolidated fund management can simplify it.

5. Maintain a balance in the state general fund.

Some states protect themselves against erroneous revenue estimates and downturns in collections with provisions that prohibit appropriating all of the expected general revenue for the coming budget period. If conditions are favorable, the revenue saved from one year can be appropriated in the next. The provisions in Delaware, Mississippi, and Rhode Island allow the appropriations of 98 percent of their general fund revenue estimate. Iowa allows 99 percent to be appropriated. Oklahoma, traditionally plagued by unpredictable revenue fluctuations because of energy prices, allows the appropriation of only 95 percent of the forecast. When the full amount of the estimate is collected, state practices differ. Revenue above the ceiling for appropriations can appropriated in a later legislative session or may be deposited in the state budget stabilization fund.

6. Create and fund a budget stabilization (rainy day) fund.

The principle of a budget stabilization fund is that a state government saves money in prosperous years for use in bad years. Forty-four states have some sort of stabilization fund, reflecting widespread enthusiasm for the idea of saving for a rainy day.

Experience over the past decade indicates that such funds are unlikely to bail out a state in serious fiscal difficulty. At the end of fiscal year 1989, rainy day funds held their highest amount ever--$4 billion--but that was the equivalent of only 1.6 percent of state general fund expenditures for that year. Five states held over half the total amount. As that suggests, most states have always held very low balance in their rainy day funds. Most states quickl7y exhausted their reserves when state revenue collections fell off in fiscal years 1990 and 1991.

Even though rainy day funds cannot solve a major fiscal crisis like those that accompanied the recessions of 1983 and 1990, they can serve as a stop-gap measure to buy time. Rainy day funds can let decision makers measure the intensity of a problem and evaluate public sentiment on how it should be addressed (spending cuts? tax increases?). An automatic deposit mechanism can reduce the political difficulty of actually putting money into the fund.

Managing the State's Finances

Balancing the budget is at the heart of state fiscal management. Every state except Vermont has a legal requirement for a balanced budget. The specific nature of the requirement varies from state to state, and it is possible in some states to roll a deficit forward from one year to the next, to employ short-term borrowing, or to use accounting practices to resolve a budget shortfall. But even when fiscal conditions are as harsh as they were in the early 1990s, states almost invariably balance their budgets. Partly this is because of constitutional and statutory balanced-budget requirements, partly because legislators or even the voters have to approve general obligation debt in most states, and mostly because the state political expectation is that the budget will be balanced annually.

Balanced budget requirements refer to the state operations budget and not to the capital budget. Unlike the federal government, state governments budget separately operations and capital expenditures. Operations includes all ongoing activities--for example, salaries, payments to vendors for medical care, rent of buildings, purchases of supplies and equipment, and grants to local governments. Capital budgets include new construction or the acquisition of land and other major equipment purchases and repairs or modifications of structures. The specific guidelines vary from state to state. In FY 21992, the 50 states' capital expenditures came to $50 billion, while states spent $561 billion on current operations and aid to local governments (note 12).

State governments increasingly borrow money to finance capital expenditures, an as fiscal conditions have tightened in recent years, the practice of financing capital projects from annual revenues has become less common. Long-term borrowing to finance operations is rare, and in recent times has occurred only in exceptional circumstances.

Balancing a budget is not a simple matter. It involves having good revenue and expenditure forecasts and mechanisms to finance or reduce state spending when revenues fail to meet expectations. This chapter discusses ways of meeting those needs.

Agree on the basic budget numbers.

Some numbers that are basic to budget negotiations can be agreed upon through establish processes, removing potentially divisive issues from debate. The following suggestions are based upon what state governments have found to work effectively.

1. Establish a constitutional or statutory process to produce a revenue forecast that is binding upon the legislature and governor.

Constitutional requirements for revenue forecasts that are binding upon the legislature and the governor have worked so well for many states that other states ought to consider following their example.

Two thirds of the states have mechanisms to produce a binding revenue forecast. As a rule, binding forecasts are produced by executive branch agencies or by a consensus process; the latter includes legislators or legislative staff. No state legislature by itself produces a revenue estimate that is binding on the executive. The advantage of a binding forecast is that is removes a difficult technical issue from political debate, or, at least, terminates the debate at some point so that all parties can move on. For the mechanism to work, the process of arriving at a binding forecast must be trusted by the governor, legislators, and the general public. The procedure has to be clear, open, and consistent from year to year.

Eighteen states have formal processes to produce a consensus forecast, requiring the agreement of at least the executive and legislative branches. Several states include a state university faculty economist or economists from the private sector in the process. In addition to the expertise academic and private sector participants add to the process, their involvement makes the revenue forecast less of a routine activity of elected or appointed officials, and can add to its credibility. Credibility also can be enhanced by an open and publicized process, which in itself can strengthen the binding character of the forecast.

Forecasts are developed from past trends, and are susceptible to error in times of rapid change. States need to supplement their forecasting system with an opportunity for occasional revision. The team that establishes the forecast should have the power to review it formally and to revise it--but not too often. Doing so more than quarterly is unnecessary since frequent revisions give too much emphasis to trends that might be short-lived.

2. Establish a process to forecast the statistics that drive the budget.

Examples of statistics that drive the budget are school enrollments, prison populations, social services agency caseloads, and the size of the state work force. Other important data of budgetary significance are long-term public obligations represented by public employee pension and health insurance plans, capital needs, and bonded debt.

The goal is to remove from debate what can reasonably be settled by technical analysis. Whether the prison population is too high is a matter for policy discussions, but the size of the prison population should be a fact people can agree on. Such numbers should be readily available to legislators, legislative staff, and the public. The executive branch is the appropriate source of such numbers, but he process of generating them should be open to legislative participation and critique, since legislative deliberations rely upon such numbers. Some states are now designing systems to give legislators and staff on-line access to agency data.

3. Establish a process to define the budget base and agree on the amount.

The most usual method of state budgeting is to begin with the current level of spending on operations (the budget base), assume the level of services it provides is the appropriate level, and then add to or subtract from the base in relatively small amounts (increments). Even as some states move toward move emphasis on performance-based budgeting--focusing on the delivery of services rather than on the amount of resources given to an agency--the base is an essential starting point.

Calculating the base can be complicated because it does not include all state spending. The base excludes capital expenditures and other one-time expenses. It may need adjustments to reflect the full annual cost of a program that was initiated in mid-year. Because the base is the beginning point for the next budget, disagreements over what it should include can slow the process.

A particular area of difficulty has to do with development of a continuation level budget. This term refers to a budget in which the base for the coming budget period is not just the amount spent in the current budget period, but where the base is increased to cover the effects of inflation, cost increases other than inflation, allowances for caseload growth, increased insurance costs, statutory cost-of-living adjustments, and any other additional amounts required to continue current levels of services. A continuation level budget makes the assumption that such costs will be covered. These budgets have the advantage of showing how much additional money is needed just to continue an existing program. But if the total amount required to cover "continuation" costs is larger than revenue growth from year to year, there can be an apparent budget shortfall even while revenues are growing.

Naturally this can be confusing to legislators and the public. Continuation budget costs can absorb year-to-year revenue growth unless growth is substantially above inflation, possibly turning apparently revenue growth into apparent shortfall as happen in some states in the early 1990s. What looks like revenue growth when only the amounts of money are considered turns into a shortfall when only the growth in costs is added. Continuation budgets easily become political footballs; a theoretically useful tool can become an impediment to the budget process.

For those reasons it is essential that budget documents specify the nature of the base they are built on and the nature of any adjustments. Amounts added to previous levels of funding in order to cover costs or caseload increases should be identified as such, and the rational for the estimates should be specified. Laying out the facts will not bring agreement with them, but can at least clarify what the numbers are intended to mean.

Review tax expenditures with the rest of the state budget.

A tax expenditures is a tax reduction or exemption designed to encourage some particular activity, for example, tax breaks to encourage economic development or charitable contributions. Such tax expenditures are the fiscal equivalent of a state expenditures for the same purpose, but only the beneficiaries of tax expenditures tend to remember their existence. They are not subjected to the examination that formal budget items regularly encounter. Comprehensive budget review should include them, and individual tax expenditures should be reviewed just as other state expenditures are.

Develop clear guidelines for capital budgeting.

States spend between 8 percent and 10 percent of their budgets on capital projects--over $50 billion in FY 1992. To make informed decisions, legislators need an up-to-date inventory of capital stock and its conditions, a long-range plan outlining future needs, and information on how capital budget requests fit into overall budget plans. Capital requests should appear in a single budget separate from operations requests to facilitate comparison of projects with each other and with long-term plans and statements of needs. Capital budget requests should include forecasts of the costs of operating and maintaining facilities as well as the costs of acquisition or construction. An orderly process of developing and financing capital facilities can assist legislators in making optimal use of resources and in using capital to foster economic and social development (note 13).

Provide a mechanism for the reduction of expenditures when revenues fall short.

Reducing planned spending after a budget has been passed is painful and politically difficult. There is no such thing as a painless cut. A formal process for making reductions can mitigate the political difficulty by clarifying the roles of the governor and the legislature. Additionally, for the majority of states where legislatures are not in session year-round, an established process can permit a timely response to a sudden problem.

Well-defined roles for governors and legislators are more important than having the roles fit any particular model; the powers that legislatures and governors have to cut budgets differ greatly from state to state, and from time to time (note 14). A few years ago, the California Legislature removed the governor's former power to cut state budgets when there was a revenue shortfall. In June 1993, Minnesota legislators delegated increased power to their governor to cut the budget in 1993 and 1994, if necessary. Alabama and F legislators were somewhat chagrined in 19991 and 1992 to learn that their traditional practices of allowing the governor to make cuts were unconstitutional; that meant that the legislature had to share the onus of cutting the budget in mid-year.

Fifteen states give their governors full authority to cut the budget when there is a revenue shortfall. Very few states prohibit the governor from making any spending cuts. California, which as a full-time legislature, is exceptional in doing so. Vermont delegates budget-cutting authority to an eight-member Joint Fiscal Committee when the legislature is not in session.

Generally, state constitutions give governors limited authority to make cuts and require legislative participation when the extent of gubernatorial authority is inadequate. Maryland, for example, allows the governor to cut any line item appropriation by as much as 25 percent. Connecticut and Kentucky limit such cuts to 5 percent.

Whether or nor legislators are involved in budget cuts after budgets are enacted is an issue to be settled on the basis of state constitutions, traditions, and preferences. But there should be a specific mechanism in place before the need for it arises.

Beware of quick fixes that create long-term problems.

Circumstances alter cases, so that sound policy for one situation may be unsound policy when circumstances change. Various legislatures have done the opposite of what is recommended below because they saw no alternative. They will do so in the future for the same reason. But these are old rules that it makes sense to observe when possible:

1. Avoid using one-time revenues for continuing expenditures.

2. Calculate the full annual cost of a program that begins operations in the middle of a fiscal year.

3. Don't look at the state retirement system as a savings account for state government.

Be very cautious about changing assumptions about future developments connected with a state retirement system, especially if the change would reduce the current state contribution to the system.

4. Don't expect too much from early retirement incentives.

Early retirement incentive programs can be expensive in the long run if they are not carefully designed and subjected to an actuary's calculations of costs. Where savings exist, they are most likely to result from a reduction in the number of state employees. Such reductions can be difficult to preserve. If not carefully managed, early retirement incentive programs can substantially increase a pension fund's long-term liability.

5. Protect the tax base through good and bad times alike.

When prosperity makes revenue collections higher than expected, protect the tax base with rebates rather than tax cuts. Years of exceptionally high collections do not last, and cutting taxes at a time of economic expansion can lead to demands for tax increases when economic growth slows. When recessions make collections smaller than expected, make any necessary tax increases temporary. This can help avoid the cycle of higher revenues and expanded spending when prosperity returns.

Making and Communicating Decisions

Total state budgets range from about $1.5 billion in South Dakota to nearly $100 billion in California. South Dakota's Legislatures has 12 weeks to consider its budget; the California Legislature has about 20. In both cases--and in most other states--consideration of the legislative budget occurs in the midst of all the other legislative activities of a regular session. So compressed and vast a process can confuse legislators and the governor, let alone the taxpayers who make it all possible and for whose benefit budgeting is supposed to occur.

Every state's budget process has unique features. Legislatures vary greatly in their independence of the executive branch, in the proportion of legislators who participate directly in the process, in the amount of time it takes to write a budget, in the attention they give to budget administration and program evaluation, and in partisanship. A fair and open process does not depend upon any specific set of procedures and there's no way to ensure a timely, smooth completion.

The points that follow are not intended as recommendations that every state should adopt. States are too distinctive for any formula to apply. Instead, this chapter is intended to draw attention to the varieties of solutions legislatures have found for problems that they have in common.

The budget cycle--annual and biennial budgets

One of the most common recommendations for budget reform is to change the budget cycle: to a biennial budget in an annual-budget state, and vice versa. The tide has long run in favor of annual budgeting. About half the states have shifted from a biennial to an annual budget over the past 50 years.

The tide may be turning. Connecticut and Nebraska returned to biennial budgeting in recent years, and such a change has been recommended for the federal government. Current reformers often assume that biennial budgeting is superior to annual budgeting because of the greater opportunity a biennial cycle provides for planning and evaluation. Vice President Al Gore's National Performance Review, for example, endorses a national biennial budget to eliminate "busy work" that prevents "evaluating programs and meeting customer needs" (note 15).

The arguments for and against annual budgeting are well known and have not changed for many years. They are discussed at length in a separate NCSL report, "Annual and Biennial Budgeting: The Experience of State Governments." The gist of that report can be summarized with two quotations it contains. The first is from a report on budgeting cycles that the Council of State Governments released in 1972:

In reality, a state can develop a good system of executive and legislative fiscal and program planning and controls under either an annual or a biennial budget. The system would work differently with the alternative timespans, but could be effective under either approach (note 16).

The Public Affairs Council of Louisiana looked at the same issue 10 years later and reported:

The arguments used to justify and refute both annual and biennial budgets remain essentially unchanged [since the CSG report 10 years before] and unproven. The success of a budget cycle seems to depend on the commitment of state officials to good implementation rather than on the method itself (note 17).

Biennial budgeting is sometimes said to be more conducive to long-term planning, program review, and evaluation than annual budgeting because more time is available.

Long-term planning. Evidence from states that have changed from annual to biennial budgeting over the past 30 years fails to provide strong support for the contention that biennial budgeting is conducive to long-term planning. Studies from such diverse sources as the Council of State Governments, Texas A&M University faculty, and the United States General Accounting Office similarly fail to reach a definite conclusion on either side of the issue (note 18).

Program review and evaluation: An attractive argument for biennial budgeting is that is allows more time for performance evaluation, and thus can encourage administrators and legislators to focus on budget outcomes instead of controls. This was one of the principal arguments that led Nebraska and Connecticut to return to biennial budgeting in recent years. Connecticut proponents contended that "the present system (of annual budget) does not allow enough time to review expenditures in depth. Those preparing the budget finish one year and then immediately plunge into the next year's budget" (note 19). Legislators and staff in Nebraska have made the same point, and say they have indeed used the extra time for more extensive program review and evaluation.

Biennial budgeting thus may offer the opportunities for program review and evaluation that are essential to current proposals for budget reform. Leaders in the Texas Legislature, which has gone further than any other state in adopting performance budgeting, contend that the biennial budget cycle was essential to their reforms. Robert Junell, chair of the Texas House Appropriations Committee, has commented that "even though we had over 20 years' experience with performance data, we could never have made as much progress in redesigning the Texas system had we not had the luxury of a biennial interim." It provides time for the development of long-range plans, performance measures, and benchmarks, and for all the negotiating among legislators, executive branch officials, and agency staff the process required.

Since oversight of performance budgeting requires a lasting commitment of time and attention from legislators, biennial budgeting probably will do more than allow the luxury of start-up time. The same commitment of time will be easier to meet in a biennial-budget state than in an annual-budget state.

Beginning the process early

1. Begin legislative hearings before the governor has submitted a budget.

In most states, the legislators wait for the governor's budget recommendations before starting budget deliberations, although in some states agency budget planners submit their budget proposals to legislators (or legislative staff) at the same time that the proposals go to the governor. This lets legislators and their staffs begin work on the budget as early as September or October of the year before they have to vote on a budget. An early start allows for greater understanding and consideration of agency proposals and requests. It also allows for more program review and evaluation than may be possible in budget hearings during a legislative session.

Except for states with year-round sessions, such a schedule requires a commitment of substantial legislative time when the legislature is unlikely to be in session. Part-time legislatures may find it very difficult. It also requires legislative access to agency budget requests and personnel, which some governors oppose. Solving these problems enriches the process in terms of increased legislative knowledge of agencies and budgets, increased opportunity for executive officers to make their case to the legislature, and increased public knowledge of state finances through hearings and media coverage.

2. Encourage the governor to submit the budget early.

A governor's budget is the most important comprehensive planning document must legislators ever see. Even in the handful of states like Texas and Colorado where the enacted budget is in all senses a legislative product, the governor's budget sets the agenda for a legislative session as no other document can. When governors issue their budgets late, they damage the state policy making processes by truncating the time available for consideration. Governors should recognize the importance of a timely budget submission and observe statutory and constitutional requirements.

Opening the budget process to more legislators

How much rank and file legislators are involved in the legislative budget process varies greatly among the states. In some states, committees are small and powerful, and on-members may feel shut out of the budget process because their direct participation is limited to floor votes. In other states, a small, powerful committee has proved satisfactory. A prime example is Colorado where the powerful Joint Budget Committee includes only six of the 100 legislators. The even more powerful Joint Finance Committee of the Wisconsin Legislature--the unique example in the United States of a joint budget and tax committee--is limited to 16 members out of the 132 Wisconsin legislators.

At the opposite end of the spectrum from Colorado and Wisconsin, legislatures in other states are using two different ways to bring a very large proportion of legislators directly into the budget process:

• The legislatures in Iowa, North Carolina, and Utah appoint between two-thirds and all of the legislature to budget subcommittees. Legislators who are not on the budget committees are on revenue committees, so that every legislator has a direct, formal role in budget legislation. This does not mean that every legislator has an equal role in deciding fiscal issues. Such broad participation has required strong leadership roles in each of the three states, and majority rule still prevails.
• In the Minnesota Legislature, budget consideration has been moved to policy committees. "Spending divisions" within the policy committees generate the appropriations bills. The House Ways and Means Committee and the Senate Finance Committee reconcile the appropriations bills to produce a balanced budget. All legislators serve on at least one spending division or on a revenue committee, so that all legislators are directly involved in the fiscal process.

Other ways legislatures open the process include:

1. Reasonably large budget and finance committees. These may make use of small subcommittees for in-depth study of particular agency budgets or proposals; they may include minority party members on committees in proportion to their strength in a chamber.
2. An accessible budget bill or bills. In states that use one long budget bill, legislators who are not on the budget committee can be faced with the review of a long and complicated bill all at once. If one bill is desirable, a way to make it more accessible is to release it piecemeal for informal review before reporting it from committee. The goal is to prevent obstacles to legislators' understanding of what they are expected to vote on.
3. Comprehensive and timely summaries of appropriations bills.
4. Rules requiring enough time to elapse between the reporting of budget legislation from committee and floor votes to let members read through the bill or bills.
5. Coordination of the work of budgeting and policymaking committees through the use of fiscal notes and fiscal impact studies.

The role of leadership

Legislative leaders (plus the governor in some states) have the essential roles of maintaining a budget schedule and reconciling all budget proposals to produce a balanced budget. This is essential to legislative budgeting, not just a matter of leaders demonstrating their power. The challenge for leaders is to make the necessity of this role clear to other legislators and to the public.

Budgets have to be balanced and ought to be completed on schedule. Someone has to pull it all together. Legislatures like those in Iowa, North Carolina, and Utah, where most or all legislators are on budget-writing subcommittees, still require the leadership to make major decisions. In each case, the leadership of the two chambers agree on budget targets for the budget subcommittees before the process begins. At the end of the process, the leaders reconcile the subcommittee recommendations to produce a balanced budget.

This essential role cannot be carried out by subcommittees or committees because of the difficulty of reaching agreement. It has to be centralized.

Communication with the public

It is important to have clear and definite information on budget decisions and state financial management with different levels of detail available to the public.

1. Summarize the budget in readily understandable documents and make bill summaries and fiscal impact studies readily available.

Short, clear summaries of budget policy and budget decisions are essential for communicating with the media and with voters generally. Some good examples are Cal Facts: California's Economy and Budget in Perspective, published annually by the Legislative Analyst's Office in Sacramento; the Fact Book published each year by the Finance, Ways and Means Committees of the Tennessee General Assembly; the Michigan Senate Fiscal Agency's annual Statistical Report; the Summary of Recommendations and Economic and Revenue Forecast that the Arizona Joint Legislative Budget Committee publishes annually; and the Legislative Handbook Series published by the Maryland General Assembly.

Tax study groups, other public interest groups, and private sector organizations sometimes distribute such summaries. The work of the Rhode Island Public Expenditure Council is an outstanding example (note 20).

Where such studies are not readily available, legislators should take measures to have them produced. They are an invaluable means of reaching the public.

2. Require a Comprehensive Annual Financial Report (CAFR) that meets the standards of the Government Finance Officers Association (GFOA).

The GFOA annually presents certificates of achievement to governments that publish readable, efficiently organization, comprehensive annual financial statements that also comply with generally accepted accounting principles. Examples abound--Minnesota's CAFR has won the award for a number of years and is a very good example of clarity of organization and presentation (note 21).

Such documents are of fundamental importance in providing voters, financial institutions, and other governments with timely, reliable information. They also contribute to building confidence in government.

Conclusion

Writing a budget for a state government involves the most complicated and controversial issue in public life: how the public's money gets spent. Given the number and variety of interests and issues that have to be reconciled for a budget to be completed, the wonder is that the process moves along as smoothly as it does year after year. But for many observers, it is the competitiveness, compromises, and incomplete nature of the process that are striking, not the real accomplishment every annual and biennial budget represents.

This report has presented some suggestions for easing the process at specific points. It has reported on solutions some legislatures have found, or experimented with, to problems that most legislatures face. There's no assurance that what works in one state will work in another; state government exists in this country to let individualism flourish, and in budgeting as in many other areas of state government, there are as many solutions to problems as 50 groups of creative individuals can invent. The point of this report is not that there is one tidy solution to every problem legislators see in their budgeting process. The point is that there is a great deal of inventiveness to be seen in how states have grappled with the process, and that there is always an alternative to an unsatisfactory process. The NCSL fiscal partners who put this report together hope that it is an encouragement to legislators, staff, and everyone concerned with the state budget process to keep on looking for better ways to write budgets, to keep on experimenting, and to keep on learning from their neighbors.

Notes

1. Mulloch v. Maryland, 4 Wheaton 316, 431.
2. Hal Hovey, "Many Faces of Budget Reform Tempting to Policymakers," The Fiscal Letter (National Conference of State Legislatures) 15, no. 2 (March/April 1993).
3. This discussion of performance budgeting draws substantially upon Karen Carter, The Performance Budget Revisited: A Report on State Budget Reform, Legislative Finance Paper no. 91 (Denver, Colo: National Conference of State Legislatures, 1994).
4. Allen Schick, Zero Base '80: The Status of Zero-base budgeting in the States (Washington, D.C.: National Association of State Budget Officers and the Urban Institute, 1979), p. 2
5. Schick, Zero Base '80, passim.
6. Aaron Wildavsky, The New Politics of the Budgetary Process, 2nd ed. (New York: Harper Collins Publishers, 1992), p. 426.
7. Hovey, "Many Faces of Budget Reform."
8. Marcia A. Howard, State Tax and Expenditure Limits: There is No Story (Washington, D.C.: National Association of State Budget Officers, 1988); Philip G. Joyce and Daniel R. Mullins, "The Changing Fiscal Structure of the State and Local Public Sector: The Impact of Tax and Expenditure Limitations," Public Administration Review 51, no. 3 (May/June 1991): 240-253.
9. Iris Lav, presentation to the National Conference of State Legislatures workshop "The Tax Revolt after Ten Years: Dead or Dormant?" Tulsa, Okla., August 10, 1989.
10. See Steven D. Gold, The State Fiscal Agenda for the 1990s (Denver, Colo.: National Conference of State Legislatures, 1990), pp. 131-148.
11. Advisory Commission on Intergovernmental Relations, Changing Public Attitudes on Governments and Taxes: 1993. (Washington, D.C.: ACIR, 1994).
12. U.S. Bureau of the Census, State Government Finances, 1992 (Washington, D.C., U.S. Government Printing Office, 1993), table 1.
13. Barbara Yondorf and Barbara Puls, Capital Budgeting and Finance: The Legislative Role (Denver, Colo.: National Conference of State Legislatures, 1987).
14. Corina L. Eckl, Legislative Authority Over the Enacted Budget (Denver, Colo.: Fiscal Affairs and Oversight Committee, National Conference of State Legislatures, 1992).
15. Albert Gore, Creating a Government That Works Better & Costs Less, Report of the National Performance Review (Washington, D.C.: U.S. Government Printing Office, 1993), p. 17.
16. Council of State Governments, Annual or Biennial Budgets? (Lexington, Ky.: CSG, 1972), p. 23.
17. Public Affairs Research Council of Louisiana, "Results of PAR Survey on Annual vs. Biennial State Budgeting" (Baton Rouge, La., 1982).
18. CSG, Annual or Biennial Budgets, 1972, p. 23; Public Affairs Research Council of Louisiana, "Results of PAR Survey"; Charles W. Wiggins and Keith E. Hamm, Annual Versus Biennial Budgeting? Public Policy Paper no. 7 (Austin, Texas: Public Policy Resources Laboratory, Texas A&M University, 1984), III-15; United States General Accounting Office, Budget Issues: Current Status and Recent Trends of State Biennial and Annual Budgeting (Washington, D.C., 1987).
19. [Connecticut] Commission to Study the Management of State Government. Final Implementation Report (Hartford, Conn., 1991).
20. Rhode Island Public Expenditure Council, 300 Richmond Street, Providence, R.I. 02903.
21. The Minnesota Comprehensive Annual Financial Report is prepared by the Department of Finance, 400 Centennial Building, 658 Cedar Street, Minneapolis, Minn. 55155.

Acknowledgments

This report was the work of many contributors: legislators, legislative staff, and representatives of members of the Foundation for State Legislatures. Besides drawing upon the expertise of those who participated directly, the NCSL staff who compiled this report made use of many published documents from state legislatures and academics, some of which are noted in the footnotes. Jerry Sohns, director of development for the Foundation for State Legislatures, and Corina Eckl were the principal NCSL contributors to this report, and they acknowledge the assistance of Karen Carter, Scott Mackey, Arturo Pérez, and Judy Zelio of the Fiscal Program staff. Michelle Pavlock prepared the text for publication, and it was edited by Patricia Wunnicke.

Participating Legislators and Legislative Staff

Joyce Bigbee, Director, Legislative Fiscal Office, Alabama
Representative Nancy Brown, Kansas
Thomas L. Covington, Director, Fiscal Research Division, North Carolina
Bill Goodman, Assistant Director, Bureau of Legislative Research, Arkansas
Representative Kathleen W. Gurnsey, Idaho
Robert Keaton, Director, Senate Office of Fiscal Affairs and Policy Development, Louisiana
James Jimenez, Legislative Finance Committee, New Mexico
Delegate Nancy Kopp, Maryland
Brian Maddy, Chief of Staff, Oklahoma State Senate
Representative Guy R. Nadeau, Maine
Paul Reuss, Secretary, Senate Finance Committee, New York
Representative Paul Schauer, Colorado
Robert Sumrall, Budget Officer, House Appropriations Committee, Mississippi

Representatives of the Foundation Fiscal Partners

American Express Company

Les Goldberg, Vice President, State Government Affairs
Stephen D. Lemson, Director, State Government Affairs

American Federation of State, County, and Municipal Employees

Ann Kempski, Department of Public Policy
Marie Monradd, Assistant Director for Public Policy

AT&T

Linda Adams, Government Affairs Director, Idaho, Montana, and Wyoming

Eastman Kodak Company

Stephen J. Ciccone, Manager, State Government Relations
Judith D. Peteres, Director, State and Local Government Relations

National Education Association

Joseph Falzon, Professional Associate, Economics and School Finance, Research Division
Education Hurley, Research Specialist
R. Dick VanderWoude, Senior Professional Associate, Government Relations

PepsiCo

Bill Ehrig, Government Affairs
Galen Reser, Director of Government Affairs
David L. Wright, Vice President, Government Affairs

Philip Morris U.S.A.

Derek Crawford, Government Affairs
Patrick Riscitelli, Issues Manager, Public Affairs
Tina Walls, Director of Government Affairs

Published June 1995. Page reviewed January 2004.
Email mailto:statebudget-info@ncsl.org?subject=Fundamentals of Sound State Budgeting Practices for more information.
Visitor counts for this page.

Contents

• The Foundation Fiscal Partners
• Fundamentals of a Sound State Budgeting System
• Acknowledgments
• 1. The Challenge to Traditional State Budgeting
• 2. Constraints on State Government
• 3. Managing the State's Finances
• 4. Making and Communicating Decisions
• Conclusion