Statement of
Douglas Holtz-Eakin
Director

Reforming the Federal Budget Process

before the
Subcommittee on Legislative and Budget Process
Committee on Rules
U.S. House of Representatives

March 23, 2004

Chairman Pryce, Congresswoman Slaughter, and Members of the Subcommittee, thank you for the opportunity to testify on the issue of reforming the budget process.

The long-term outlook for the federal budget is daunting. Over the coming decades, the aging and retirement of the baby-boom generation will place extraordinary pressures on federal programs for the elderly--principally Social Security, Medicare, and Medicaid. As lawmakers face the difficult policy choices that will be needed to address that trend and other pressing budget issues, it is useful to evaluate the fundamental structure of the budget process within which those choices will be presented and made. At a minimum, the budget process should present budgetary choices in a clear and transparent fashion, should be an effective framework for reaching and carrying out budget policies, and should include procedures to ensure that those policies are enforced.

I will make the following points:

• The existing budget process provides the means for lawmakers to establish and enforce comprehensive budget policies and works best when an underlying consensus on fiscal goals has been achieved.
  
  
• The statutory budget disciplines that expired in 2002--the limits on discretionary spending and the pay-as-you-go (PAYGO) requirement for new legislation affecting entitlements and revenues--proved to be an effective enhancement of the budget process. As with the underlying process, however, they worked more effectively when supporting broad agreement on multiyear fiscal goals. When such agreement did not exist, the disciplines were largely ignored.
  
  
• Identifying the long-term obligations imposed by new entitlements or expansions to existing entitlements is a valuable objective of some proposals, but how to account for those obligations in the budget is a complicated issue. More work is needed before a full system of budgeting and enforcement for such obligations can be implemented.
  
  
• Various proposals intended to restructure the process more broadly or impose stronger spending discipline--such as a joint budget resolution, biennial budgeting, caps on mandatory spending, and enhanced rescission--might have some benefits but also might not be as powerful as envisioned and could have unintended consequences, such as a transfer of power to the executive branch and increased complexity in the budget process.

The Scope of the Current Budget Process

The federal budget process is an amalgam of procedures that lawmakers and public officials use to plan, establish, control, and account for spending and revenue policies. In general, the budget process includes the preparation of the President's budget by the executive branch, the Congressional budget process (centered on a Congressional budget resolution and, in some years, on reconciliation legislation), the authorization and appropriation process, the execution of budget law (including impoundment, a procedure under the Congressional Budget Act of 1974 for deferring or rescinding appropriated funds), and financial management.

Under the Budget and Accounting Act of 1921, the President submits a proposed budget on the first Monday in February. Under the Congressional Budget Act of 1974, the Congress's first major action is to adopt the annual budget resolution, which is scheduled to occur by April 15. It serves as a blueprint for Congressional action on separate pieces of spending and revenue legislation. In addition, the resolution's aggregate levels of spending and revenues and spending allocations made to Congressional committees are enforced by points of order that Members of Congress may raise against individual spending or revenue bills as they are considered by the House or Senate. If a point of order raised under the Congressional Budget Act is sustained, the legislation may not be considered further. The budget resolution may also instruct Congressional committees to produce reconciliation legislation that conforms permanent spending or revenue laws within their jurisdiction to levels set forth in the resolution.

The existing budget process thereby provides the means for lawmakers to establish and enforce major changes in budget policies. Over the years, the process has served as a framework for major policy initiatives and multiyear budget agreements, which often have been put in place in legislation developed to carry out reconciliation directives in budget resolutions. However, when a consensus on such policies has not emerged, the process has stalled.

Although the budget process includes enforcement procedures, it is generally neutral regarding particular outcomes. For example, there is no generic requirement that the budget resolution recommend a balanced budget. However, from 1986 through 2002, lawmakers supplemented the budget process with a temporary framework of statutory deficit control procedures that established specific goals for reducing projected deficits, capping spending, or ensuring that new legislation would be budget-neutral. Some lawmakers suggest that such goals should be made a part of the budget process for an extended period, or permanently, in light of the long-term budgetary pressures that are looming.
 

Statutory Measures to Enhance Budget Discipline

By the mid-1980s, the issue of persistent budget deficits had come to dominate the annual budget debate. In response, lawmakers enacted a new framework of statutory constraints intended to reduce and eliminate those deficits. The Balanced Budget and Emergency Deficit Control Act of 1985 (known as the Gramm-Rudman-Hollings Act) established a schedule of fixed, declining deficit targets for every fiscal year beginning in 1986 and leading to a target of zero in 1991. The Deficit Control Act also created the procedure of sequestration to automatically cut spending for many federal programs if the deficit for a fiscal year was estimated to exceed the target level. A sequestration, if necessary, would be carried out by an executive order issued under the terms of a sequestration report from the Comptroller General of the United States, the head of the General Accounting Office. That report was to be based on a joint report by the Office of Management and Budget (OMB) and the Congressional Budget Office (CBO).

In 1986, the Supreme Court held in Bowsher v. Synar that having the President's sequestration order, an executive action, be determined by a report from the Comptroller General, an official accountable to the Congress, was unconstitutional. Therefore, in 1987, the Deficit Control Act was modified to give the OMB Director sole authority to prepare the estimates and calculations used to trigger a sequestration order. As part of that change, CBO was required to issue advisory sequestration reports. The 1987 modification of the law also revised the deficit targets and extended them through 1993.

Although deficits shrank somewhat in the late 1980s (from $221 billion in 1986 to $153 billion in 1989), they failed to meet the statutory targets--in some years by substantial margins. The targets set under the Deficit Control Act, both the original ones and the revised ones, were unrealistic in light of the prevailing economic conditions. For that reason and others, actual deficits remained above the targets during the years that the law was in effect.

An Overview of the Budget Enforcement Act

To strengthen the budget process, the Budget Enforcement Act (BEA) was enacted in the fall of 1990 as an amendment to the Deficit Control Act. The BEA was part of a multiyear agreement to reduce deficits that was embodied in the Omnibus Budget Reconciliation Act of 1990 as title XIII. Representing a different philosophy of deficit control, the BEA established procedures to ensure that the deficit reductions enacted in the 1990 budget agreement would be carried out. With the BEA, lawmakers set rules that would hold them accountable for changes in the deficit due to new legislation--but not for the budgetary effects of economic and other factors outside of their immediate control, the factors that played the most significant role in the difficulties in meeting the Deficit Control Act's targets.

The BEA established a budget enforcement framework that divided the budget into two parts. Discretionary spending, which is provided and controlled by appropriation acts, would be subject to annual aggregate limits on budget authority and outlays. Laws affecting mandatory spending and revenues would be covered by a PAYGO procedure to prevent those laws from increasing the deficit. A breach of the discretionary spending caps would lead to reductions only in discretionary programs, and a breach of the PAYGO control would trigger cuts only in certain mandatory programs.

Originally set to expire at the end of fiscal year 1995, the discretionary spending limits and PAYGO requirement were amended and extended twice, in 1993 and again in 1997, as a part of two subsequent multiyear deficit-reduction agreements. In each extension, the basic framework of the BEA was continued without major substantive changes. With the emergence of surpluses in 1998, some people asserted that the PAYGO requirement should be interpreted to apply in a fiscal year only if new mandatory spending or revenue laws were estimated to cause a deficit to return. However, both OMB and CBO, with the concurrence of the House and Senate Budget Committees, continued to prepare PAYGO estimates and sequestration calculations without regard to estimates of the deficit or surplus for a particular fiscal year. At the end of fiscal year 2002, the discretionary spending limits and PAYGO requirement expired.

The Effectiveness of the BEA

During most of the period that the BEA procedures were in place, the federal government's fiscal fortunes improved significantly. Deficits declined steadily after 1992, and beginning in 1998, surpluses were recorded every year through 2001. The BEA framework contributed to that turnaround. Between 1991 and 1997, total discretionary outlays were nearly level in nominal dollars (see Figure 1), although there were substantial shifts in spending--aided by the end of the Cold War--from national defense to nondefense programs. During that same period, most new mandatory spending and revenue laws were consistent with the PAYGO requirement to be deficit-neutral; end-of-session balances on the PAYGO scorecard consistently showed zero or net reductions in the deficit.
 

Figure 1.

Actual Discretionary Outlays Compared with the Spending Limits as Originally Enacted

(Billions of dollars)

Source: Congressional Budget Office.

But the surplus in 1998, the first recorded in nearly 30 years, eliminated the essential purpose of the BEA--to combat and control deficits. In that new fiscal landscape, with projections showing mounting surpluses for the coming decade, the BEA could not restrain the pressures to spend more.

To comply with the letter of the law while boosting discretionary spending above the statutory limits, lawmakers used a number of approaches--including advance appropriations, delays in making obligations and payments, emergency designations, and specific directives. For example, in 1999 and 2000, lawmakers enacted emergency appropriations totaling $34 billion and $44 billion, respectively--far above the annual average for such spending from 1991 to 1998. Comparable amounts were enacted for 2001 and 2002 mainly in response to the terrorist attacks of September 11, 2001. Similarly, from 1991 to 1997, emergency appropriations totaled $52 billion (excluding amounts provided for the Persian Gulf War in 1991 and 1992), but during the four years following the 1998 surplus, emergency appropriations totaled more than three times that amount.

To accommodate increased nonemergency spending for 2001, lawmakers raised the caps on budget authority and outlays by $99 billion and $59 billion, respectively. The following year, they increased the limits on budget authority and outlays by even larger amounts--$134 billion and $133 billion, respectively. From 1998 through 2002, total discretionary appropriations (in nominal dollars) grew at an average annual rate of 8.5 percent; by comparison, from 1991 through 1997, such spending declined at an average annual rate of 1.1 percent.

Similarly, after the emergence of surpluses, lawmakers enacted legislation to increase mandatory spending or reduce revenues but used legislative directives to statutorily comply with the PAYGO requirement. Thus, for 2001 and later years, lawmakers eliminated more than $700 billion in positive balances--that is, amounts that would have triggered a PAYGO sequestration--from the scorecard (see Table 1). Most of that amount stemmed from the estimated drop in revenues attributed to the Economic Growth and Tax Relief Reconciliation Act of 2001. By contrast, during the earlier years of the BEA, the balances on the scorecard were zero or negative, and lawmakers statutorily removed negative balances so that those savings could not be used to offset the costs of new mandatory spending or revenue legislation.

Table 1.
Balances Eliminated by Statute from the Pay-As-You-Go-Scorecard
(Billions of dollars)
	1997	1998	1999	2000	2001	2002	2003	2004	2005	2006	Total, 1997- 2006
Eliminated Balance	-9	-3	0	-3	90	65	127	150	142	144	701
Source: Congressional Budget Office using data from the Office of Management and Budget's final sequestration reports, fiscal years 1991 to 2003.
Note: Positive numbers indicate an increase in the deficit or reduction in the surplus; that is, eliminating positive balances removed the need for a PAYGO sequestration. Negative numbers indicate a decrease in the deficit or increase in the surplus; that is, eliminating such balances made them unavailable to be used as an offset to additional mandatory spending or a reduction in revenues.

The Impact of Sequestration

During the 12 years that the threat of a discretionary sequestration was present, sequestrations were ordered only twice, both in 1991 (the first year that the spending limits were in effect) and both for relatively insignificant amounts. One of the sequestrations was rescinded by subsequent law; the second led to estimated savings of $1.4 million (discretionary spending totaled $533 billion in 1991). For laws affecting mandatory spending or revenues, a PAYGO sequestration has never been triggered.

Interpreting the absence of large sequestrations over the BEA's history is difficult. In some years, especially 1991 to 1997, the threat of sequestration perhaps served as an effective deterrent to legislation that would have violated the spending limits or PAYGO requirement. From 1998 to 2002, the absence of sequestrations may simply have reflected the lack of consensus among lawmakers to guard the bottom line of the budget. With the emergence of large surpluses came the willingness to enact legislation to increase the caps substantially or eliminate the positive PAYGO balances. The lack of sequestrations may also have reflected shifting priorities; for example, legislative efforts aimed at reviving the economy and fighting the war on terrorism were deemed more important than avoiding a return to budget deficits. Such a change in priorities may confirm that a budget enforcement framework works best when a consensus exists on the fiscal goals to be achieved and the policies needed to achieve them.

Issues in Reinstating the BEA

Despite recent experience, the underlying philosophy of the BEA--that enforcement procedures should target the budgetary effects of new legislation--has proved to be effective when there was a political consensus to reduce deficits.

As lawmakers decide whether or how to reinstate those procedures, they may want to be mindful of some of the problems that occurred. The occasion of considering proposals to reimpose caps on discretionary spending and the PAYGO requirement also could be an opportunity for a broader reexamination of how the previous process could be improved. Some issues include the following:

• Budget "Firewalls" for Discretionary Spending. In some years, lawmakers created separate caps for spending on defense, domestic programs, international programs, transportation, victims of crime, and conservation. By design, that approach reduces the flexibility of the appropriation process by setting priorities in advance. However, because spending priorities may shift from year to year, if the overall caps are extended for a five-year period--as they have been in the past--establishing sublimits might make it difficult to shift priorities while constraining total discretionary spending.
  
  
• Emergency Spending. The BEA included an exemption for any spending or revenue legislation declared to be an emergency requirement by the President and the Congress. The use of the emergency exemption, especially after 1998, led lawmakers and others to question whether much emergency spending was for true emergencies or was simply a way to appropriate more funds under tight discretionary caps without having to find offsets. The BEA exemption for emergency spending required only that the President and the Congress both agree on the amounts to be designated; it did not limit those amounts or restrict the purposes for which they could be provided.
  
  Some observers feel that the emergency exemption should be replaced with a system of budgeting for emergency needs that is based on an annual average amount of emergency spending appropriated in previous years. Others would strictly limit the amount of funding that could be designated as an emergency requirement or would establish a strict statutory definition of emergencies to guide legislative action on such spending. Those approaches could also be combined. However it is fashioned, an emergency safety-valve procedure of some type that allows additional resources to be provided for unexpected contingencies is probably essential to allow the government to address unanticipated and pressing needs. Too much flexibility, however, could render caps on discretionary spending ineffective.
  
  
• Inflation Adjustment to the Discretionary Caps. Until 1997, the BEA stipulated that the caps on discretionary spending were to be adjusted for changes in the rate of inflation from that anticipated when the caps were originally established. Although inflation has been low in recent years, restoring an inflation adjustment may help to sustain political agreement on cap levels over a longer period.
  
  
• Sequestration. The effectiveness of sequestration has been questioned. That only two small sequestrations have been ordered, that caps on discretionary spending have been adjusted or increased by large amounts, and that large PAYGO balances that would have triggered a sequestration have been eliminated by law all point to potentially significant limitations in the procedure.
  
  If the sequestration procedure is to be resurrected, one issue that lawmakers would need to address is the exemption of most mandatory programs from a PAYGO sequestration. Under the expired PAYGO process, if a sequestration was triggered, the total amount of spending available to cut, because of specific exemptions and special rules, was quite limited. Thus, the brunt of any sequestration would have fallen on a relatively small pool of mandatory spending. For fiscal year 2003, CBO estimated that less than $65 billion in mandatory outlays would have been subject to a PAYGO sequestration, an amount that was less than 5 percent of total mandatory outlays for the year.

Long-Term Obligations and the Budget Process

The President and some Members have proposed adding new procedures to the budget process to focus attention on the long-term costs of entitlement programs. At a minimum, such proposals would require the Office of Management and Budget and the Congressional Budget Office to prepare annual analyses of the long-term imbalance, or the present value of the excess of expenditures over dedicated receipts, for Social Security, Medicare, and federal civilian and military retirement and health programs over at least 75 years. Gradually, more programs could be added for analysis. In addition, the President would be required to estimate the effect on the long-term fiscal balance of any legislation included in his budgetary proposals. For its part, CBO would be required to estimate the effects of reported legislation on the long-term balance of any applicable group of entitlement programs.

That information could be a supplement to cost estimates, like that provided under the Unfunded Mandates Reform Act. Alternatively, new procedures could create points of order against legislation that would worsen the current imbalance or against legislation lacking such an estimate. Neither alternative would require any change in the budgetary treatment of entitlement programs but would serve to identify legislation that had a significant impact on mandatory outlays.

While those proposals are innovative and worthy of further study, the proposed measure of fiscal imbalance, which nets projected dedicated revenues from expenditures, deserves careful attention. It might be more useful to require separate treatment of expenditures and dedicated receipts, both for analysis and points of order. For instance, a point of order might be raised against legislation that would either increase the present value of expenditures or reduce the present value of dedicated receipts. That change could make programs' costs more transparent and reduce the potential for strategic manipulation of the imbalance measure. However, cost estimates based just on the net present value of future cash flows might mask important information on the timing of those transactions that lawmakers could find useful.
 

Broader Reforms of the Budget Process

Recent experience with the budget process has caused frustration among some lawmakers, who have raised doubts about the effectiveness of simply reinstating the BEA procedures. They contemplate broader reforms that are intended to make the process more efficient and effective, in addition to improving fiscal discipline. Those proposals include these:

• Convert to a Biennial Budget Cycle. Proposals for biennial budgeting generally call for policymakers to enact budget legislation one year and to oversee and evaluate activities in the next. Supporters of biennial budgeting are increasingly concerned that the requirements of the annual budget process are overwhelming policymakers and public officials. They argue that the seemingly incessant demands of that process detract from other functions of government--such as long-range planning and oversight--that are equally, if not more, important.
  
  However, changing to a two-year cycle could have significant drawbacks. It could diminish the effectiveness of Congressional control of spending in the appropriation process and could make adjusting to rapidly changing budgetary and economic conditions more difficult. If the economic and technical assumptions underlying a two-year budget resolution were not revised before the end of the biennium, the information and estimates that policymakers used would be far less reliable. Budget discipline and Congressional oversight might suffer if major policies were reviewed less frequently. And large supplemental appropriation bills could be more likely under a biennial budget process. Finally, it is unclear whether the root cause of the problems cited by proponents of biennial budgeting is the annual timetable or other factors that would be largely unaffected by a switch to two-year budgeting. Slightly less than half of the states, which are cited by proponents as a proving ground for biennial budgeting, use a biennial budget cycle. Many of the larger states tend to budget annually.
  
  
• Make the Budget Resolution a Law. Each year, the President and the Congress propose separate budget plans. When those plans differ fundamentally, final agreement on spending and tax legislation is difficult to reach. The President and the Congress could be required to make the budget resolution a law each year.
  
  On the one hand, enacting the budget resolution could promote earlier agreement on priorities between the President and the Congress. In addition, a statutory budget resolution might be a more effective means to pair new budget policies with the appropriate enforcement procedures, such as discretionary caps and a PAYGO requirement. Combining budget policies and enforcement procedures in that manner also could be a better way to ensure that current enforcement procedures reflected lawmakers' most recent consensus. On the other hand, a statutory resolution might not make overall agreement on the budget easier, and if the requirement delayed the resolution, Congressional action on regular appropriation bills and on revenue or other spending legislation could become stalled as well. Converting the budget resolution into a law also would fundamentally alter the Congressional budget process, as the process would no longer be one for setting Congressional priorities but would focus on the joint priorities of the President and the Congress. The change could have broader effects and could lead to a shift in budgetary power to the President.
  
  
• Adopt Controls on Mandatory Spending. Since the 1960s, outlays for entitlements--such as Social Security, Medicare, and Medicaid--and other mandatory spending programs have risen substantially as a share of the economy. If current policies remain unchanged, that trend is projected to continue, with mandatory spending (excluding offsetting receipts) rising from 11.8 percent of gross domestic product in 2004 to 13.1 percent in 2014. And long-term budgetary pressures caused by the aging of the baby-boom generation will only exacerbate that trend.
  
  As a result, some observers advocate caps on mandatory spending enforced by sequestration, patterned after the caps on discretionary spending. Total mandatory spending could be capped at levels that permitted a limited rate of growth, and any spending over that level would automatically result in an across-the-board cut. However, such an approach would be difficult to implement because such spending depends on a variety of factors that vary significantly from year to year and because automatic reductions in such programs might affect beneficiaries in ways that would be politically unacceptable to the Congress. For example, under the Medicare program, a sustainable-growth-rate formula has been established to control spending for physicians' services, but in recent years lawmakers enacted legislation permitting the applicable spending target and payment rates to be increased rather than allowing scheduled reductions to take effect. And, more generally, if a significant amount of mandatory spending was exempted from sequestration, as it was under the BEA's PAYGO requirement, a comprehensive cap on mandatory spending might be ineffective or could distribute the burden of enforcement unequally among federal programs.
  
  
• Establish a Mechanism Like the Line-Item Veto--Expedited Rescission or Separate Enrollment. The Supreme Court invalidated the Line Item Veto Act in 1998. That law, enacted in 1996, set in place a procedure for the President to cancel certain provisions providing targeted tax benefits or spending that he deemed wasteful or unnecessary. But the Court held that the procedure violated the presentment clause of the Constitution.⁽¹⁾ Since then, at least two alternatives have been introduced in the Congress that supporters hope will revive the budget control device in a constitutional fashion. The first, expedited rescission, would ensure that the Congress voted on the President's proposed cancellations. The other, separate enrollment, would require each tax benefit or spending "item" in a bill passed by the Congress to be presented separately for the President's approval.
  
  Spending control disciplines similar to the line-item veto continue to attract interest because they are viewed as a way to control "pork barrel" spending. However, whether such procedures would save significant sums or would simply shift spending priorities to those favored by the President is unclear.

Conclusion

The federal budget process should provide clear and transparent information on budgetary choices, provide lawmakers with a framework for developing a consensus on spending and revenue policies, and enable those policies to be enforced. The Budget Enforcement Act proved to be helpful during most of the 1990s in moving the nation toward a balanced budget, at least in part because there was a political consensus that supported that objective. In the current environment, some variant of that system could be worthwhile. In general, proposals to change the budget process should be judged on the extent to which they promote those fundamental purposes and help policymakers to confront the impending challenges of the long-term outlook.