Hearings of the Committee on Rules
"Biennial Budgeting: A Tool for Improving Government Fiscal Management and Oversight"
Ronald Snell
Economic and Fiscal Division Director of the National Conference of State Legislatures
1. When you say that "there is little evidence of clear advantages of either annual or biennial state budgeting practices," you are stating that there are no clearly inherent advantages to either cycle as found in the states. Therefore, is it not the case that neither side of this debate can truly use state experiences to settle this debate? Secondly, while the state experience cannot answer the fundamental question, can we not examine each of the cycles in practice in the various forms of the 50 states and find clues as to how to actually implement a biennial process at the Federal level? What are some of those clues with respect to preparing, considering and monitoring budgets?
You noted, first, that my testimony indicated that "there are no clearly inherent advantages to either cycle as found in the states" and asked "is it not the case that neither side of this debate can truly use state experience to settle this debate?" In my mind, state experience demonstrates that the length of the budget cycle is only one, and probably not the significant, element in determining the success of a state's budget process. I think that partisanship, fiscal circumstances, and policy issues are equally important at any given time. Any state's budget process, annual or biennial, may run to a smooth conclusion in one budget session and be hesitant, quarrelsome, divisive and unsatisfactory in another budget session, depending on how those various issues interact.
Moreover, federal and state budgeting differ in ways that keep state experience from being a decisive predictor of federal experience. The important differences (besides those of scale) include:
- The role of federal taxation and spending in the national economy—no state budget, including the budgets of the largest states, can decisively affect the national economy;
- The greater separation of powers at the federal level, which may be contrasted to the continuous inter-action of the legislative and executive branches that characterizes many states' budget and appropriations processes;
- The state emphasis upon balancing its general-fund resources and expenditures within each budget period, which is less emphasized by the federal government;
- Separate state capital budgets; and
- The lack of separate authorization and appropriations processes in state government (of which more below in response to your fourth question).
You also asked, "can we not examine each of the cycles in practices in the various forms of the 50 states and find clues as to how to actually implement a biennial process at the Federal level?"
My colleagues at NCSL and I would draw the following state practices to your attention as being of possible value to Congress in reconsidering its budget processes. These do not necessarily characterize either annual or biennial budget states, but they are state practices that merit your consideration.
A. A balanced budget. Although fewer than half the states have a constitutional requirement to enact a balanced operating budget each year or biennium, practice and tradition make balanced operating budgets a primary concern in every state. (Some states are allowed constitutionally to fund capital projects through borrowing which is not included in the balanced budget equation). The general expectation that operating budgets will be balanced against available revenues has benefited fiscal discipline in the states. As a matter of formal policy, NCSL recommends a balanced federal budget except in times of national emergency.
B. Legislative program evaluation and oversight. States employ a great variety of assessment tools, ranging from agency involvement in the legislative appropriations and budget process through traditional performance audits to highly sophisticated systems of strategic planning and performance evaluation. While only Florida and Texas have made much progress in linking performance to budgetary allocation decisions, many states have enriched legislators' knowledge of program activities and performance through such processes. The National Conference of State Legislatures and the Urban Institute are engaged in a multi-year study of state programs of governing-by-results. The final report of this study is scheduled for release later in 2000 and I look forward to sharing it with you.
Although NCSL has not taken a formal position on this point, in my view one of the strongest arguments for a biennial federal budget is the opportunity it would provide for Congress to adapt the authorization process to allow greater legislative oversight of program performance. The Government Performance and Results Act of 1993 lays the foundation for such oversight in its requirement that federal agencies develop strategic plans and goals and track their progress toward those goals with specific performance measures. Congress has unparalleled resources in the GAO, the CRS and the CBO to assist in independent assessment of agency operations. Freedom from the pressures of annual budgeting could allow Members greater opportunity to make use of such information in program review.
C. Greater Legislative-Executive interaction. The state budget process is characterized by more frequent interaction of the legislative and executive branches than the federal process. While state processes vary greatly, a frequent practice is for executive agencies to submit their budget requests to the legislature at the same time the requests go to the state executive budget office. State legislative committees can hold hearings on agency requests at the same time the executive branch is preparing its comprehensive budget recommendations. Such a process vastly enriches legislators' understanding of agency budgets and requests and facilitates final consideration of the budget. It also facilitates budget negotiation between the branches early in the process. Congress, of course, has to be more alert to the separation of powers than state governments, as a rule, but it might be possible to adapt this process to Congressional use.
D. A consensus revenue estimating process. The states have been steadily moving toward a consensus revenue estimating process, through which the two legislative chambers and the executive branch reach agreement on a revenue forecast. The process usually involves an elected official or staff person from each chamber and the governor's office and may include academic or private sector economists or policy analysts. In some states constitutional provisions make the estimate final and binding, and in other states periodic revisions are permitted. Whatever the details of the process, the goal is to remove gamesmanship from the revenue estimating process, and to settle the issue early in the budget cycle.
E. A separate capital budget. Thirty-three states budget separately for capital needs and operational needs. States do so partly because they increasingly use debt to finance capital expenditure and do not do so to finance operating expenses. However, the distinction is a useful one in encouraging separate consideration of long-term public needs from continuing operations.
2. You mentioned that many states have other procedural mechanisms that they utilize to assist in fulfilling or maintaining their budget processes regardless of whether they are on an annual or a biennial cycle. What have states used to deal with unanticipated contingencies and what can the Federal government learn from this use?
Your second question addressed state mechanisms for dealing with unanticipated contingencies.
The 42 states whose legislatures do not meet year-round have devised various ways to address the needs for budget reductions, transfer of funds among or within agencies and appropriation of unanticipated revenues when the legislature is not in session. I have attached tables from the NCSL publication Legislative Budget Procedures (1998) that show the processes state-by-state.
3. You state that "according to older studies of state decisions to shift to annual budgeting from biennial budgeting, supplemental appropriations became less common after the shift. But in recent years supplemental appropriations have been common in all states - not just those with biennial budgets - because of the unpredictable changes in the national economy and because of cost overruns in Medicaid programs." Could you elaborate on this point?
Your third question addressed the point made in my written testimony that supplemental appropriations had become more common in all states in recent years.
Because 46 state fiscal years run from July 1 to June 30, legislatures that convene in January begin their work at a time that is suitable for reviewing the status of the budget. States adjust their budgets in the middle of fiscal years for three principal reasons: to address revenue shortfalls within balanced-budget constraints, to address revenue collections above original projections and to address needs that were unforeseen when the budget was originally enacted. Examples of the last are enrollments in elementary and secondary schools (which are a factor in determining state aid to school districts and which cannot be known with precision until after the school year begins), corrections inmate population changes, and changes in Medicaid enrollments and other income-supplemental programs.
The precise reasons for supplemental appropriations and other mid-course budget adjustments change with time. The need to adjust appropriations for public school enrollments continues in most years. In the early 1990s, state revenue collections tended to fall below estimates and inmate and social-service populations increased above projections, requiring supplemental appropriations for those programs. In some states it appeared that inmate and social-service population growth had purposefully been under-estimated at the time the original budget was written in order to facilitate a balanced budget.
The adverse economic circumstances that produced those results (and strategies) have been reversed. In recent years, supplemental appropriations have been less necessary for social service programs, and have been more at legislative discretion as a way to make use of unanticipated revenues.
4. Do states differentiate between authorization and appropriation legislation as does the Federal government and how does this affect the way state legislatures conduct programmatic oversight?
Your fourth question asked whether states differentiate between authorization and appropriation legislation as the federal government does, and asks how the answer to that question affects state legislatures' programmatic oversight.
The federal authorization process has no parallel in state government. State programs are created in legislation and the enabling legislation remains codified in statute, as is the federal practice, but state enabling legislation does not routinely authorize a maximum or specific level of program expenditure when the program is created or subsequently. That is true for all 50 states.
The legislative appropriations process provides appropriations of state funds for agency use with the general understanding that the full amount of the appropriation is to be expended for the purposes stated. Some states practice program budgeting (Tennessee and Louisiana are good examples) and attach few conditions to expenditures within a program. Most states use fairly detailed line-item appropriations, and the current interest in program-performance budgeting has only slightly decreased the practice.
Programmatic oversight occurs in a number of ways that differ among the states.
A. In 32 states, legislatures receive agency budget requests from the agencies before the executive budget in prepared. The general practice is for legislative budget committees to review the requests (either in hearings or through legislative staff analysis) before the legislature convenes. In hearings, legislators can question agency staff directly about their budget requests. Staff analyses can be used to raise questions about the use of previous appropriations as well as about requests for increases or changes in budgets.
B. Legislators use budget hearings to question agency staff on existing programs and past budgets as well as budget requests. Governors' budget proposals have tended to present a growing amount of explanatory and performance data, which encourages legislative questions on program efficiency and effectiveness.
C. Thirty-three states have enacted or are in the process of enacting performance measurement and performance reporting processes of varying stringency, with the goal of increasing legislative oversight of agency efficiency and effectiveness. About 15 additional states have created similar programs by executive order. NCSL and the Urban Institute are studying these processes with the goal of releasing major reports on them in the fall of 2000, which we would be happy to share with you.
D. State auditors (who are legislative officials in more than half the states) carry out performance audits as well as compliance and financial audits. Where auditors are legislative officials, such audits are chosen by legislative committees. Kansas and Minnesota offer excellent examples of the use of this process as a tool of legislative oversight. Legislative fiscal staff also conduct performance reviews in some states.
