Hearings of the
Committee on Rules
"Biennial Budgeting: A Tool for Improving Government Fiscal Management and Oversight"
Susan J. Irving
Associate Director, Federal Budget Issues
Accounting and Information Management Division
1. In your written testimony, you assert the importance of taking a longer view on matters involving fiscal management of the federal government. However, you note that there is no single, formal mechanism that helps to focus that "longer view." Is such a formal mechanism needed, in your view? What sort of mechanism would you argue is needed? What would it look like and how would it be enforced?
I don't really think that a new mechanism is needed for improving information about the longer-term budgetary consequences of various program and budget decisions. There are several ways that a longer view could be incorporated into budget decision-making under the current structure and process. The use of accrual budgeting for certain programs, like federal insurance and employee pensions in which the current cash-based budget measure does not adequately represent the extent of the government's commitment to make future payments, would lengthen the time horizon considered in budgeting decisions for those programs.(1) Increasing and possibly formalizing the use of modeling to simulate the fiscal and economic outcomes associated with alternative policy paths would inject a greater consideration of the long term and the effects of Social Security and health care. Mechanisms to incorporate consideration of such simulations might include points of order when certain growth rates are exceeded. Or if growth in a particular program exceeds expected growth by some specified percentage, the Congress could be required to vote to accept the growth or offer some alternative. Alternatively, the President could be required to initiate alternatives to address unexpected growth or to declare his support for accepting the growth.
We have suggested that Congress consider whether a new mechanism is needed to facilitate cross-cutting oversight. In testimony before the Senate Budget Committee in February 2000,(2) the Comptroller General noted that while individual authorization and oversight committees are well-suited to address performance issues affecting individual agencies or programs, many of the key performance questions are not confined to, and cannot be addressed effectively on, an agency-by-agency or committee-by-committee basis. Many areas are addressed by a wide range of mandatory and discretionary spending programs, tax expenditures, and regulatory approaches that cut across federal agencies and committee jurisdictions. The President is now required to prepare and submit to Congress a governmentwide performance plan including fiscal, management, and program performance expectations; however, Congress has no established mechanism to articulate performance goals, assess strategies to achieve these goals, or to define an oversight agenda based on crosscutting performance and management issues. One possible approach would be to modify the current budget resolution to include a "performance resolution" to permit the Congress to respond to, and present a coordinated congressional perspective on, the President's governmentwide performance plan. This might involve collecting the "views and estimates" of authorization and appropriation committees on priority performance issues, and working with the Governmental Affairs Committee and others to identify crosscutting performance concerns. This approach would be one mechanism to assist Congress in routinely using program performance information to enhance oversight.
2. Your testimony takes a close look at state experiences with biennial budgeting. Do you find this type of focus on the states to be relevant to the federal government's experience with budgeting? Critics of biennial budgets have argued that there are too many differences between the federal and state levels of government to make a proper comparison between the two. Based upon your research, what is your view?
I strongly agree with the proposition that extreme care should be used in extrapolating from state budgeting practices and processes to the federal level. As I noted in my statement, state budgets fill a different role, may be sensitive to different outside pressures, and otherwise are not directly comparable to the federal budget. Nevertheless, we can learn from them. What processes and institutions have been developed in the states where biennial budgeting is seen as successful? How well would those translate to the federal level? What would be the impact on the balance of power between the Congress and the executive branch? Given the particular interest here in using a biennial cycle to strengthen oversight, it is worth looking at what the states do in that area—again, with the view that the mechanisms adopted by the states offer models or cautions for the federal government.
3. You allude in your testimony to the importance of considering the impact of a change to biennial budgeting on the relative balance of power. Could you give the committee your view of how this balance would be impacted should biennial budgeting (including appropriations) be implemented?
Annual enacted appropriations have been, and remain, a basic means of exerting and enforcing congressional policy. When agencies come before you to request funding, they expect to answer questions about how they have followed congressional intent. If the appropriations cycle is shifted from annual to biennial, what will replace this kind of oversight? The relative balance of power between branches is not dependent solely on any single process. Rather, it is the cumulative result of many processes and practices. My point was that a shift to a biennial appropriations cycle will change one set of congressional-executive interactions. In implementing such a change, attention should be paid to the cumulative effect of small changes on the relative balance of power.
4. A major argument in support of biennial budgeting is that it would free up the Congress to conduct more thorough oversight of federal programs and agencies. You touch on this point in your testimony. Could you elaborate on the current state of affairs with regard to the level of oversight that occurs in Congress under the existing annual budgeting system? What is your view of how a biennial process would impact oversight? What reasons would you give for the perennial lack of authorization for certain programs (current CBO assessment lists approximately 130 programs totaling more than $120 billion that lack authorization)?
Oversight occurs in many ways and in many venues. I spoke of the need for a more global view of federal activities and an assessment of the way the government seeks to achieve its goals and objectives, a type of oversight that may not even fit within the current committee structure. Under the current structure, appropriations subcommittees provide oversight when they question agencies about planned and past performance. In fact, some of the strongest oversight can occur when funding is at stake. Whether a biennial process would increase oversight, and by whom, depends on how it is implemented. If appropriations were actually held to a biennial cycle with few "off-year" legislated changes, the appropriations subcommittees' members would spend less time on appropriations—but since appropriations have long served as a vehicle for agency- or program-specific oversight, it is unclear how this would increase oversight in this forum. Since it is unclear what in the annual appropriations process is thought to hinder oversight by authorization committees, it is unclear how a shift to a biennial cycle would affect it.
It seems difficult to attribute what you refer to as the "perennial lack of authorization for some programs" to the annual budget cycle. Although we have not made a systematic study of this area, where appropriations are made without an authorization, the situation seems to reflect an agreement among the relevant players to proceed in this fashion. It is not clear that a change in the budget cycle would change the underlying issues that lead to a decision to proceed this way.
5. Do you subscribe to the argument that, without annual action on agency budgets, Congress would have ceded some of its power and would end up having less direct control over those agencies and programs?
The appropriations process is a compelling venue for agency oversight but it is not the only one. As I stated in my testimony, the shift to a biennial budget process will bring neither the end of congressional control nor the guarantee of improved oversight. Clearly a change in the appropriations cycle would change the nature of congressional oversight and control. Whether it would lessen that control would depend on how the change is implemented.
6. You discuss the DoD experience with biennial budget submissions. Some critics of biennial budgeting point to this example as evidence that biennial budgeting will not reduce redundancy or save effort. Is it possible to evaluate this partial experiment with biennial budgeting, given the fact (as you point out in your statement) that there have not been biennial authorization or appropriations for DoD?
The requirement to submit biennial budgets has caused additional workload for DoD because it must prepare estimates and justifications for the second year of a biennium twice—once when submitted as part of a biennial budget and again the next year since biennial appropriations have not been provided. We cannot really draw conclusions from the implementation of the DoD requirement because of Congressional reluctance to make biennial authorizations and appropriations. This reluctance would need to be overcome in order to implement biennial budgeting.
7. Could you address the concern that, under a biennial budget structure, agencies would end up padding their budget requests to ensure adequate funding for the second year? How might this be impacted (or offset) by an agency's ability to obtain efficiency savings through longer planning horizons for management and procurement?
It is difficult today for agencies to project funding needs for a fiscal year that will end 20 months after the budget is submitted. This will be even more difficult if the projection period is increased by 12 months. As a prudent manager, I suppose an increase in the time horizon over which I was asked to project funding needs would lead me to wish to round up. However, as long as there are discretionary spending caps, there is little room for what you refer to as "padding." I expect that there would be greater uncertainty about the second-year estimates than about the first-year estimates. You may need to institute an adjustment process to compensate for such uncertainty—although if adjustments become too widespread, then it is an annual cycle again. Agencies with substantial procurement activity generally have multi-year or no-year authority; thus many of the benefits of a longer planning horizon are already available.
8. In your testimony you mentioned that since the mid-1960's, 17 states have changed their budget cycles. What has been the trend in these changes since 1990?
During the 1990s only two states—Arizona and Connecticut—have changed their budget cycles. Both moved from annual to biennial: Arizona in 1999 and Connecticut in 1993.
9. Can you elaborate on the differences and similarities between multi-year appropriations, forward funding and biennial appropriations?
A multi-year appropriation is budget authority that is available for a specific period of time in excess of one year, typically for two, three, or five years. Forward funding is budget authority that generally becomes available for obligation in the last quarter of the fiscal year and the authority continues into the following fiscal year. Thus, it is budget authority which covers a period of time that does not coincide with the federal fiscal year. A biennial appropriations process, as defined in pending legislation, would provide funding for two one-year appropriations cycles. Multi-year authority and forward funding are two of the ways the Congress has shown its willingness to accommodate program cycles that are either different from or longer than a single fiscal year. Advocates view the shift to biennial appropriations as another way to ease budget execution by providing more certainty in funding over 2 years.
10. As you know, the current budget process only accounts for changes in mandatory spending and revenues that result from legislative action. OMB and CBO hold the budget harmless for changes in spending or revenue levels caused by administrative, economic, or technical changes. Do you think it would be possible and advisable to utilize this existing mechanism (or a modification thereof) as a way of handling such changes in the budget under a biennial process?
Limiting the amount of "off-year" appropriation actions will be achieved only if congressional and executive leaders are committed to making it happen. While it might be attractive to apply the "pay-as-you-go" concept to off-year changes under a biennial process, I think it would present problems, especially if the off-year changes were for emergency or unanticipated events—exactly the types of changes that most agree would be acceptable under biennial budgeting. For example, would PAYGO apply to legislation enacted in the off year in the aggregate so that enactment of emergency legislation must be offset by enactment of some discretionary spending reductions at some point during the session of Congress? This could result in additional appropriations actions in what was to have been the year devoted to other congressional activities. Or would each piece of legislation be expected to contain its own offsets so that it is cost-neutral? This would possibly limit the number of new appropriations but may be difficult to achieve if the new appropriation is truly for an emergency. Would off-year appropriations be able to be offset by changes in the direct spending/receipts part of the budget, such as by revenue increases, or would the offset have to be from discretionary spending? The questions are complicated. We would be happy to sit down and work through the implication of various alternative approaches.
A. How long of a ramp up period do you think is necessary to ensure that both the executive and legislative branches are ready to implement this new process?
B. Specifically, how long will it take agencies to obtain the ability to prepare, propose, and manage two-year budgets?
C. Would you suggest that biennial budgeting be phased in over some set time period with certain portions of the budget moving to a biennial schedule one year followed by others? If not, would you advocate not moving any budget submissions to a biennial schedule until the entire budget is ready to proceed in that direction? If so, would the Department of Defense be the most likely first candidate for such a transition?
Although I would defer to my colleagues at OMB and CBO who are more directly affected, it seems reasonable that it would take two or more years to fully implement this change because of adjustments to systems, processes, and guidance to agencies. Another consideration in timing the transition would be to avoid switching to a 2-year process in the first year of a new administration. Following initial implementation Congress may want to consider whether the calendar in the first year of a biennium is reasonable. Current proposals shorten the time available for appropriations action in the first session of the Congress – when all appropriations action is supposed to take place – by reducing the time between completion of congressional action on the budget resolution and the beginning of the biennium at issue by 1 month.
As I stated in an earlier question I would anticipate some degree of uncertainty or difficulty in developing estimates because of the longer time horizon between when these estimates are developed and when appropriations are enacted. Regarding budget execution, most agencies of any size are already managing multiple appropriations accounts that provide budget authority for various periods so this transition should not be difficult. Even agencies with annual appropriations should not have problems adjusting as long as budget authority is made available in two 1-year increments.
In general, I do not believe that there is any advantage in an incremental transition. Identifying stable agencies likely will be difficult – what may appear stable at a department or agency level masks greater volatility at the appropriations account level-- and I suspect that some candidates that are relatively stable may not be good test cases because of size. Having said this, DoD would be the department best positioned to go first. The mechanisms are in place and it does receive the largest portion of discretionary funds.
12. What role would you envision reconciliation playing under a biennial process specifically in the off-year? I would expect that the reconciliation process would also become biennial when the Congress decides to use this procedure in the budget process.
