United States General Accounting Office GAO Testimony Before the Committee on Rules, House of Representatives For Release on Delivery Expected at 9:30 a.m. BUDGET PROCESS Wednesday, May 12, 1999 Comments on H.R. 853 Statement of Susan J. Irving Associate Director, Budget Issues Accounting and Information Management Division GAO/T-AIMD-99-188 Mr. Chairman and Members of the Committee: It is a pleasure to return to talk with you again about the congressional budget process--and especially whether and how it should be changed to meet the fiscal situation presented today. Attached to my testimony today is a list of testimony statements we have issued on the budget process over the past 5 years. In addition, I was fortunate to participate in your September 1997 briefing on budget enforcement procedures in the House of Representatives. 1 As we have discussed before, everyone involved in the budget process shares some frustration with it. The public finds it confusing. Executive branch agencies say it is burdensome and time-consuming. Members of the Congress say it seems too lengthy with too many votes on authorizations, budget resolutions, reconciliation, appropriations, emergency supplementals, and the debt limit. In one sense, of course, nothing could be more important than debates about the budget. Budgeting is the process by which we as a nation resolve the large number of often conflicting objectives that citizens seek to achieve through government action. The budget determines the fiscal policy stance of the government--that is, the relationship between spending and revenues. And it is through the budget process that the Congress and the President reach agreement about the areas in which the federal government will be involved and in what way. Because the decisions are so important, we expect a great deal from our budget and budget process. We want the budget to be clear and understandable. We want a process that presents the Congress and the American people with a framework in which to understand the significant choices and the information necessary to make the best informed decisions about federal tax and spending policy. In addition to these broad goals, the budget process has also been expected to respond to the budget challenges of a particular time. The 1974 Budget and Impoundment Control Act was designed to reassert the Congress’ role in setting overall federal fiscal policy and in establishing spending 1 The Congressional Budget Process committee print, September 26, 1997. Page 1 GAO/T-AIMD-99-188 priorities. The act sought not to achieve a particular outcome but to impose a structure and a timetable on the budget debate. It was neutral as to fiscal policy. It was not until the enactment of the Balanced Budget and Emergency Deficit Control Act (also known as Gramm-Rudman-Hollings or GRH) in 1985 that the budget process was designed to achieve a particular goal. Both GRH and the 1987 amendments to it sought to achieve a specific outcome: a balanced budget by a time certain. However, GRH sought to use a change in process to force agreement on substance--and measured against its stated objective of a balanced budget, it did not succeed. The 1990 Budget Enforcement Act (BEA) took a different tack toward the same end. While it also sought to achieve a balanced budget, it used process to enforce a previously reached agreement. It was designed to limit congressional actions that would increase the deficit. On its own terms BEA succeeded, but its ambition was limited. It did not seek to control economic or demographic-driven growth in existing entitlement programs--and that is the area of greatest growth today. Nevertheless, the combination of fiscal discipline and economic growth led to the first balanced budget in nearly 30 years. Today, therefore, a different fiscal situation has emerged. After nearly 30 years of unified budget deficits, current projections are for “surpluses as far as the eye can see.” At the same time, the country faces a demographic tidal wave that will--absent a change in policy--overwhelm the budget. This is a new set of challenges for the budget process: almost 30 years of projected surpluses followed by--absent changes in Social Security and Medicare--a reappearance of large and growing deficits. These circumstances present an opportune time to reexamine the budget process. Such an examination should be guided by a number of key principles. General Criteria for a In the past we have suggested four broad goals or criteria for a budget process.2 The process should Budget Process 2 Budget Process: Evolution and Challenges (GAO/T-AIMD-96-129, July 11, 1996) and Budget Process: History and Future Directions (GAO/T-AIMD-95-214, July 13, 1995). Page 2 GAO/T-AIMD-99-188 • provide information about the long-term impact of decisions while recognizing the differences between short-term forecasts, medium-term projections, and longer-term simulations; • provide information and be structured to focus on important macro trade-offs; • provide information necessary to make informed trade-offs between missions and between the different tools of government; and • be enforceable, provide for control and accountability, and be transparent. Each of these is important, and they are related--but they cannot all be maximized in a single process. Trade-offs are necessary. Today, in the context of H.R. 853, your staff asked me to focus especially on the importance of the long-term perspective, on increasing the understanding and recognition of long-term commitments and insurance commitments in the budget, and on how this relates to the need for control, accountability, and transparency. Long-Term Perspective A long-term perspective is important in the budget debate in both a macro and a micro sense. By macro I mean the nation’s economic health. The and Commitment nation’s economic future depends in large part upon today’s budget and Recognition investment decisions.3 Therefore, we believe that at the macroeconomic level, the budget should provide a long-term framework and should be grounded on a linkage of fiscal policy with the long-term economic outlook. The micro aspect of this longer-term perspective relates to those programs and activities where a longer time horizon is necessary to understand the fiscal and spending implications of commitments for specific purposes. Examples include retirement programs, Medicare, and pension insurance-- and even some discretionary programs whose design implies continued funding. Although BEA’s multiyear focus represented significant progress in this regard, planning for longer-range economic goals and looking at the costs of some commitments requires looking much further ahead. For these programs, even very rough projections may be better than ignoring the long term. 3 Budget Issues: Long-Term Fiscal Outlook (GAO/T-AIMD/OCE-98-83, February 25, 1998). Page 3 GAO/T-AIMD-99-188 Since the bill before you deals with both of these aspects of the long term, let me discuss each. Long-Term Macro Beginning in 1992, congressional leaders have requested that we provide a long-term macro perspective by modeling the implications of different Perspective fiscal policy paths for the nation’s economy over a long-term period, which has ranged from 50 to 75 years. We have periodically updated these simulations to account for changes in the fiscal and economic environment. For the last 4 years the Congressional Budget Office (CBO) has also produced long-term simulations and the President’s budget has included long-term simulations by the Office of Management and Budget (OMB). The CBO and GAO results have been quite similar. Looking at the simulations since 1992 tells the dual story of today’s fiscal challenge: (1) the outlook has improved greatly from earlier simulations and (2) looking out over the longer term, the current situation is not sustainable. In 1992, modeling a continuation of the then-current fiscal and budget policy resulted in a deficit exceeding 20 percent of gross domestic product (GDP) by the year 2020. In contrast, today’s update shows the benefits of the difficult policy choices made by the Congress and the President and of a healthy economy: in 2020, the model indicates a surplus of 1.5 percent of GDP and does not show a deficit reemerging until 2028. However, this improved outlook does not mean that the fiscal challenges facing the country have been met. In fact, the current situation is still not sustainable over the long term. Our most recent model results indicate that if current policy were continued, by 2063 federal revenue will cover only health care, Social Security, and interest spending. To continue all other spending at current policy levels would require federal borrowing and/or revenue increases. As the Comptroller General pointed out earlier this year,4 absent any policy changes, budgetary flexibility declines drastically over time and there is increasingly less room for programs for national defense, the youth, infrastructure, and law enforcement. This is true even if we assume that the entire unified budget surplus is saved and used to reduce debt (and thus interest) from current levels. We believe these simulations provide a useful perspective that is often lacking in budget debates. They tell us that the surplus is temporary. 4 See, for example, Social Security and Surpluses: GAO’s Perspective on the President’s Proposals (GAO/T-AIMD/HEHS-99-95, February 23, 1999). Page 4 GAO/T-AIMD-99-188 Perhaps more important, they alert us to the fact that even if the surplus is “saved,” we face an unsustainable outlook. These simulations also provide a context within which to look at longer-term projections for individual programs such as Social Security and Medicare. Both of these programs use trust fund financing and accounting. As a result, we get a picture of their financial outlook by looking at the trust funds--for example, we know that under the current tax and benefit structure, Social Security’s annual cash receipts will fall short of annual cash outlays in 2014 and that the Social Security Trust Fund will be insolvent in 2034. The Trustees’ report does look 75 years out. However, in analyzing Social Security and considering alternative program changes, it is a mistake to look only at the trust fund; it is important to also recognize how Social Security fits into the budget and the economy and to understand how it grows as a share of both. Although we consider these simulations--and other long-term “projections” --to provide critically important context for budget deliberations, we would also stress that they must be interpreted carefully. Given the range of uncertainty about economic changes and the response to these changes, these simulations cannot be viewed as forecasts of budgetary or economic outcomes 50 years in the future. Indeed, the dramatic improvement in the outlook over the last 7 years shows how sensitive these results are to unanticipated shifts in economic growth or to policy actions. The simulations, therefore, should be seen as illustrative of direction and magnitude given current information about demographic and budgetary trends and the functioning of the economy. In this spirit, the approach taken in H.R. 853 has much to recommend it. Requiring reports on 75-year budgetary trends for the budget as a whole can help provide the necessary long-term context. Few of the government’s commitments are truly transient. For example, embedded in numerous programs and policy decisions are long-term relationships with states and in the international community that have fundamental implications for the cost of government over time. The inclusion in the budget of OMB’s reports and comparisons between the President’s policy proposals and current law will focus more attention on the long term and on how the President would seek to address looming problems. Having a CBO report as well will permit the Congress and other observers to make comparisons with the OMB current law report, providing an independent view. Although we do not make budget projections or estimates, as long as it is useful to the Congress we will continue our work on the long term as well. Given the level of uncertainty involved in long-term modeling--and the need to be aware of how sensitive results are to different assumptions Page 5 GAO/T-AIMD-99-188 about how the economy works--it has proven useful to have several different entities develop and maintain the ability to simulate the long term. In the past, the few players in this arena have collaborated and shared techniques, data, and analyses. This has increased the confidence that can be placed in the direction and magnitude of the results. I am sure this will continue, especially if the requirements in H.R. 853 are enacted to ensure the continued efforts of OMB and CBO. Long-Term Focus at The budget was not designed to and does not provide complete information on long-term cost implications stemming from some of the government’s Micro Level Needed in commitments when they are made. We have long advocated that Budget policymakers need information on the long-term cost consequences of today’s commitments. For programs as large as Social Security and Medicare this is important both for macro policy and for resource allocation. However, it is also important to understand the long-term implications of the commitments for those programs too small to drive the long-term outlook. A budget is about the allocation of scarce resources. Such decisions reflect a number of factors including beliefs about the appropriate role of government in various areas, judgment about the likely success of a program in achieving certain goals, and the cost of a program. It is important that Members of the Congress and the President--and citizens--be able to compare program costs on a consistent basis. A budget should be structured to permit informed programmatic decision- making across a wide range of approaches--for example, insurance, credit, asset sales, capital, grants, and direct service. This is less difficult if policymakers know what the cost of a given decision will be. Although for many programs BEA’s multiyear time frame has represented great progress, there are programs and activities where a longer time horizon is necessary to understand the spending implications of the government’s commitment-- and this commitment affects future budgetary flexibility. H.R. 853’s requirement for reports on long-term budgetary trends should also be helpful in this area. While long-term information on Social Security and Medicare has been available in Trustees’ reports--and is often cited in the debate--these are not the only programs in which the government has made long-term commitments. Civilian, military, and veterans’ retirement benefits constitute another large category of the federal government’s commitments. While some have been recognized in the budget, none of the costs of civilian or military retiree health benefits are recognized in the budget as they are earned. The same is true for veterans’ pensions and Page 6 GAO/T-AIMD-99-188 benefits. As the result of new accounting standards that require its reporting, information on the long-term liabilities of these other retirement and benefit programs is now being made available in annual financial statements. This information can supplement the information included in the budget as decisionmakers consider the costs of these programs. Programs with an apparently shorter time horizon than pension and health commitments could also benefit from a longer-term perspective. As I noted above, many government programs and policies imply costs over a relatively long period of time. For some of these--e.g., pensions--long-term costs may be easy to calculate. For others, such as decisions about the nation’s role in the world or some intergovernmental commitments, costs are more difficult the estimate. Unfortunately for analysts, ease of calculation does not always correlate with importance. Federal insurance provided to individuals and businesses against a wide variety of risks is a prime example of the type of program that may carry long-term cost implications. In 1997, we reported that the current cash- based budget generally provides incomplete and misleading information on the cost and fiscal impact of federal insurance programs. 5 The use of accrual concepts, such as budgeting for the cost of the risk assumed by the government as in H.R. 853, has the potential to better inform budget choices. In our report, we supported supplemental reporting of these cost estimates in the budget, as required by H.R. 853. We believe this supplemental reporting will allow time to validate estimation methodologies and increase the users’ comfort levels with accrual estimates before considering whether to move to a more comprehensive approach of incorporating the risk-assumed estimates into the budget numbers. H.R. 853 requires that estimates of the risk assumed by the government in these programs be disclosed in the budget. It also sets fiscal year 2006 as a date certain for moving to the comprehensive approach. We recognize that setting a date for inclusion in budget numbers may well increase agencies’ attention to and efforts to develop good quality estimates. However, the bill also sunsets this provision at the end of fiscal year 2007--thereby including these numbers in the budget for only 2 years. This seems 5 Budget Issues: Budgeting for Federal Insurance Programs (GAO/AIMD-97-16, September 30, 1997). See also Budget Issues: Budgeting for Federal Insurance Programs (GAO/T-AIMD-98-147, April 23, 1998). Page 7 GAO/T-AIMD-99-188 problematic for two reasons. First, the knowledge that the numbers would only be used in the budget for 2 years could reduce the pressure to do the hard work necessary to develop good estimates. Second, changing the basis of budget numbers for only 2 years is likely to be both burdensome and confusing. This is not to say that there should be no reexamination of a change of this magnitude. Certainly it makes sense after some number of years of experience for the Congress and the President to consider progress under budgeting on a risk-assumed basis and make a decision whether to continue or not. However, our experience with credit reform-- which is easier than estimating risk-assumed costs for insurance--tells us that a 2-year trial is too short for making such a judgment . Whatever approach to implementation is finally decided upon, I must stress that the calculation of the risk-assumed costs is complex. Some programs will be better able to make the estimates than others. H.R. 853 also calls for OMB, CBO, and GAO to report on the advisability and appropriate implementation of budgeting for the risk-assumed costs. These reports should play a significant role in a final decision about when these numbers are ready to be incorporated into the budget. Role of Long-Term Incorporating a long-term perspective into the budget process advances the goals of control, accountability, and transparency. Transparency is a Perspective in complex goal. At times it demands simplicity--and I would be the first to Increasing Control, admit that incorporating long-term cost estimates is unlikely to simplify the budget process. However, transparency can also mean “no hidden costs” Accountability, and or “few surprises.” This aspect of transparency is advanced by Transparency understanding and disclosing the long-term cost implications of as much of the budget as is possible. The Congress, the President, and the taxpayers have a right to the best information possible about the cost of the future commitments that they are making. These long-term cost estimates are also important for control and accountability. The Congress and the President are best able to control the cost of a program when it is created or modified. For example, cash-based budgeting for insurance programs provides not only incorrect, but also misleading, information about the expected cost of these programs to the federal government. If these costs were available--even as rough estimates --at the time an insurance program was proposed, policymakers could consider design elements that might reduce costs. Page 8 GAO/T-AIMD-99-188 Technical Comments Since BEA’s limits on budget authority and outlays remain in effect through 2002, care must be taken in designing the relationship between BEA and on H.R. 853 any changes in the budget laws that take effect before its expiration. For example, H.R. 853 is clear in its repeal of the current requirement to adjust the spending limits for emergencies. The interaction between the existing spending limits on budget authority and/or outlays and any joint resolution on the budget vetoed by the President is less clear. BEA contains a number of different limits on budget authority and/or outlays. For fiscal years 2001and 2002, it contains budget authority and outlay limits for discretionary programs and separate outlay limits for highway and mass transit programs. However, under H.R. 853, if a joint resolution is enacted into law, it would specify subtotals of new budget authority and outlays for nondefense and defense discretionary spending, direct spending, emergencies, and other subsets of spending if deemed necessary. If the President signs the joint resolution and it is enacted, these subtotals would replace the current law’s spending limits. However, H.R. 853 contains “fall-back” procedures for expediting a concurrent resolution if the President vetoes the joint resolution. While this is a workable way for dealing with the possibility of a presidential veto, failure to enact the joint resolution on the budget means that BEA’s limits on discretionary, highway, and mass transit are still in effect. This would lead to a situation in which the concurrent resolution would contain subtotals for defense, nondefense, and emergencies while the governing law contained statutory limits on discretionary, highway, and mass transit spending. It would appear that the concurrent resolution on the budget’s subtotals for defense, nondefense, and emergencies would serve as a blueprint/guide for congressional action on spending, revenue, and debt without the force of law. Technical Issues in Approach to Insurance Budgeting: I have previously discussed insurance budgeting and our support for having the Congress encourage the development and reporting of annual risk-assumed cost estimates with the idea of moving toward a comprehensive accrual-based budgeting approach when feasible. H.R. 853 definitely moves in that direction. I do have two technical concerns. The first is related to the budget accounting for administrative costs described in the bill. Although the bill is somewhat unclear on some issues, it appears that the goal is to make the administrative costs a part of the risk-assumed cost, a feature that has been considered as desirable. The bill specifies that all funding for administrative expenses will be displayed in the program account and Page 9 GAO/T-AIMD-99-188 that the financing account will transfer to the program account the amounts necessary to pay the administrative costs. The financing account is the nonbudgetary account that accounts for all cash flows related to the insurance program, including premiums. The bill specifies that in calculating the risk-assumed cost of insurance, administrative expenses are to be subtracted from premiums. Presumably, although the bill is not clear on this point, this is the financing source that the financing account will use to pay the administrative expenses to the program account. I would be happy to work with your staff to clarify how administrative costs are to be financed. The second technical concern relates to the financing of reestimates. The bill specifies that the amount of the reestimate shall be paid from the program account to the financing account. It is silent as to whether the program account receives a permanent appropriation for the reestimate or whether some other financing source is envisioned. As you know, in credit reform a permanent appropriation was provided for reestimates, but some have raised the issue that this does not provide agencies an incentive to make good initial estimates. Again, I will be happy to discuss this issue in greater depth with your staff. GAO Report Requirement: Finally, I note that the bill would require that we study, at least every 5 years, the provisions of law that provide mandatory spending, and recommend the appropriate form of financing for activities or programs financed by such provisions of law. Current law requires this study but leaves the timing open--it must be revised “from time to time.” We have issued reports under this requirement three times since mid-1987 and have found that it requires so significant a commitment of time and staff that it constrains our ability to do other work. Therefore, I would like to talk with your staff about how to provide the information in which you are interested promptly and efficiently. Mr. Chairman, this concludes my statement. I would be happy to answer any questions that you or the Members of the Committee may have. Page 10 GAO/T-AIMD-99-188 Page 11 GAO/T-AIMD-99-188 Related GAO Testimonies Budget Issues: Budgeting for Federal Insurance Programs (GAO/T-AIMD-98-147, April 23, 1998). Budget Issues: Long-Term Fiscal Outlook (GAO/T-AIMD/OCE-98-83, February 25, 1998). Budget Process: Evolution and Challenges (GAO/T-AIMD-96-129, July 11, 1996). Budget Process: History and Future Directions (GAO/T-AIMD-95-214, July 13, 1995). Budget Process: Biennial Budgeting for the Federal Government (GAO/T-AIMD-94-112, April 28, 1994). Budget Process: Some Reforms Offer Promise (GAO/T-AIMD-94-86, March 2, 1994). (935315) Leter Page 12 GAO/T-AIMD-99-188 Ordering Information The first copy of each GAO report and testimony is free. Additional copies are $2 each. Orders should be sent to the following address, accompanied by a check or money order made out to the Superintendent of Documents, when necessary, VISA and MasterCard credit cards are accepted, also. Orders for 100 or more copies to be mailed to a single address are discounted 25 percent. Orders by mail: U.S. General Accounting Office P.O. Box 37050 Washington, DC 20013 or visit: Room 1100 700 4th St. NW (corner of 4th and G Sts. NW) U.S. General Accounting Office Washington, DC Orders may also be placed by calling (202) 512-6000 or by using fax number (202) 512-6061, or TDD (202) 512-2537. Each day, GAO issues a list of newly available reports and testimony. To receive facsimile copies of the daily list or any list from the past 30 days, please call (202) 512-6000 using a touchtone phone. A recorded menu will provide information on how to obtain these lists. For information on how to access GAO reports on the INTERNET, send an e-mail message with “info” in the body to: email@example.com or visit GAO’s World Wide Web Home Page at: http://www.gao.gov United States Bulk Rate General Accounting Office Postage & Fees Paid Washington, D.C. 20548-0001 GAO Permit No. GI00 Official Business Penalty for Private Use $300 Address Correction Requested
Budget Process: Comments on H.R. 853
Published by the Government Accountability Office on 1999-05-12.
Below is a raw (and likely hideous) rendition of the original report. (PDF)