Skip Navigation

This Congressional Budget Office (CBO) memorandum responds to a request from the Chairman of the House Committee on the Budget. CBO was asked to review the current budgetary treatment of emergency spending, highlight recent trends in emergency appropriations, and discuss various options for changing the way policymakers budget for emergencies. Portions of this memorandum are drawn from CBO testimony on the budgetary treatment of emergency spending given before the House Budget Committee's Task Force on Budget Process on June 23, 1998.

Ellen Hays and Sandy Davis of CBO's Special Studies Division prepared the memorandum under the direction of Arlene Holen. Priscilla Aycock, Chief of the Scorekeeping Unit in CBO's Budget Analysis Division, provided a tremendous amount of assistance. Sherry Snyder edited the memorandum, Liz Williams proofread it, and Judith Cromwell and L. Rae Roy prepared it for publication with help from Sharon Corbin-Jallow. Laurie Brown prepared the electronic versions for CBO's World Wide Web site.

Questions about the analysis may be addressed to Ellen Hays or Sandy Davis.
 
 

INTRODUCTION

BACKGROUND

TRENDS AND MAJOR PURPOSES

• Offsetting Emergency and Nonemergency Supplementals
• Trends in Domestic Emergency Spending
• Effects of the Discretionary Caps on Emergency Spending

OPTIONS FOR CHANGING THE BUDGETARY TREATMENT OF EMERGENCY SPENDING

• Eliminate the Emergency Exemption
• Require a Supermajority Vote for Emergencies
• Establish Criteria for Emergency Spending
• Establish a Reserve Fund for Emergencies

CONCLUSION

APPENDIX
 

TABLES
1.	Overview of Emergency Spending, 1991-1999
2.	Emergency Spending, 1991-1999
A-1.	Emergency Spending in Supplemental Appropriations, 1991-1999
A-2.	Emergency Spending in Regular Appropriations, 1991-1999
A-3.	Number of Emergencies in Supplemental Appropriations, 1991-1999
A-4.	Number of Emergencies in Regular Appropriations, 1991-1999
A-5.	Supplemental and Regular Appropriation Acts That Include Emergency Spending
FIGURES
1.	Emergency Spending, 1991-1999
2.	Supplemental Appropriations Net of Rescissions, 1985-1998
3.	Supplemental Spending for Natural Disasters (1976-1990) and Domestic Emergency Spending (1991-1999)

 

INTRODUCTION

At the end of the 105th Congress, lawmakers approved $21.4 billion in emergency spending as part of the Omnibus Consolidated and Emergency Supplemental Appropriations Act for 1999. That is the highest level of emergency spending enacted in the 1990s, excluding spending for the Persian Gulf War. The 1999 funds addressed a wide variety of purposes ranging from increased security at U.S. embassies and a continuation of peacekeeping efforts in Bosnia to aiding U.S. farmers affected by adverse weather. Many people have questioned the extent to which the $21.4 billion constitutes spending for true "emergencies."

"Emergency" is not clearly defined in budget process law. Under procedures that have been in effect since 1991, emergency spending is generally whatever the Congress and the President deem it to be. Emergency spending has been more than just the funding typically provided to the Federal Emergency Management Agency (FEMA) to respond to natural disasters such as hurricanes and floods. It has also included military funding for Operation Desert Storm in fiscal years 1991 and 1992, spending in the wake of such events as the riots in Los Angeles (1992) and the terrorist bombing in Oklahoma City (1995), and funding for peacekeeping efforts in Bosnia (1997).

What these events have in common is a degree of unpredictability. Disasters and other emergency situations can entail sudden and unexpected demands for high levels of funding. In the case of peacekeeping efforts in Bosnia or the computer conversion in anticipation of the year 2000, which both received emergency funding for 1999, the length and magnitude of the commitment can be difficult to determine in advance. In adopting the recent budget agreements and laws designed to reduce the deficit and control federal spending, both the Congress and the President have recognized the need for some sort of "safety valve" for emergencies. Under the Budget Enforcement Act of 1990 (BEA), currently in effect through 2002, spending deemed an emergency by both the Congress and the President is effectively exempt from the caps on discretionary spending and the pay-as-you-go requirement for mandatory spending or revenue legislation.

Since the BEA went into effect, some lawmakers and others have been concerned that the emergency safety valve has been used mainly to evade the BEA's constraints rather than respond to unanticipated needs. The high level of 1999 emergency spending has heightened that concern. This Congressional Budget Office (CBO) memorandum summarizes the nine-year history of emergency spending under the BEA. It reviews the evolution of the concept of the emergency safety valve, explains the budgetary treatment of emergency spending, highlights trends in emergency spending, and discusses selected options for changing the way emergency spending is budgeted.
 

BACKGROUND

Most of the accounts that typically fund emergencies are discretionary and are provided for in annual appropriation acts. They are therefore subject to the BEA's caps on discretionary spending. To keep total appropriations under the caps, the President usually requests, and the Congress usually provides in regular appropriation acts, less than what may eventually be needed for those accounts. When amounts provided in regular appropriations are inadequate to address an emergency that arises, the Congress typically will enact emergency supplemental appropriations during the fiscal year, usually at the request of the President.

Since passage of the Congressional Budget and Impoundment Control Act of 1974, a variety of attempts have been made to mitigate the budgetary effects of supplemental spending. Some efforts were directed toward better planning for supplementals. The 1974 act required that the totals in both the President's budget request and the Congressional budget resolution reflect an allowance for contingencies and "unanticipated uncontrollable expenditures" for the upcoming fiscal year--that is, amounts that might later need to be provided through supplementals. Other efforts addressed offsetting supplemental spending with rescissions. In response to passage of the Balanced Budget and Emergency Deficit Control Act of 1985, the Office of Management and Budget generally required agencies to suggest offsetting rescissions for all supplemental requests unless an agency could provide a "fully justified explanation" as to why it could not do so.⁽¹⁾

In 1987 and 1989, the President and the Congress, as part of two budget summit agreements, set caps on total appropriations.⁽²⁾ However, they did not want those caps to impede supplemental appropriations for true emergencies. The President and the Congress therefore agreed that neither branch would "initiate supplementals except in the case of a dire emergency." The agreements did not require offsets for dire emergency supplementals.⁽³⁾

In 1990, as part of a multiyear budget agreement, the Congress and the President adopted new procedures for controlling the deficit. Those procedures, embodied in the Budget Enforcement Act of 1990, established statutory limits on discretionary spending and a deficit-neutral pay-as-you-go (PAYGO) requirement for new mandatory spending and tax legislation. Unlike the "contractual" caps on spending imposed by the 1987 and 1989 budget agreements, the discretionary spending limits and the PAYGO requirement under the BEA would be enforced by automatic spending cuts, or sequestration. Those disciplines have been extended twice since 1990--most recently as part of the Balanced Budget Act of 1997--and are generally in effect through 2002.

Following the precedent set in the 1987 and 1989 budget agreements, the BEA exempts any spending designated jointly by the President and the Congress as an "emergency requirement." The discretionary spending limits are adjusted for any appropriations for designated emergencies, and the PAYGO estimates and calculations exclude any emergency mandatory spending or tax legislation.⁽⁴⁾ Totals and spending allocations under the Congressional budget resolution are also adjusted for designated emergency spending or revenue amounts.

The Congress has also provided so-called contingent emergency appropriations--appropriations that are designated by the Congress in law as an emergency requirement before they have been designated by the President. Contingent emergency appropriations are not available to be obligated until the President designates them as well. For purposes of Congressional scorekeeping, CBO scores both contingent and designated emergency spending at the time those amounts are considered by the Congress. When contingent emergency spending is designated (or accepted) by the President, whether in the same or in a subsequent fiscal year, CBO moves that emergency item to the "designated" category.
 

TRENDS AND MAJOR PURPOSES

Total emergency appropriations under the BEA have fluctuated each year from about $5 billion to $46 billion (see Table 1 and Figure 1). Those totals, however, include spending for the Persian Gulf War, which was completely offset by foreign contributions. Spending for the Gulf War (or Operations Desert Shield/Desert Storm) was a form of emergency spending under the BEA and was automatically exempt from the discretionary spending limits. It dominated total emergency spending in 1991 and, to a lesser extent, in 1992 and accounts for virtually all defense-related emergency spending in those years. Excluding spending for the Persian Gulf War, therefore, total emergency appropriations ranged from $1.5 billion (1991) to $21.6 billion (1999) and averaged about $8.9 billion annually.
 

FIGURE 1. EMERGENCY SPENDING, 1991-1999

SOURCE: Congressional Budget Office.

NOTE: Data are current as of November 19, 1998.

a. The large amounts in 1991 reflect the costs of Operations Desert Shield/Desert Storm.

After the Gulf War years (1991 and 1992), emergency spending for defense generally was intended to bolster defense readiness and fund peacekeeping missions. During that later period, it ranged from about $640 million (1993) to $7.8 billion (1999) and averaged $2.6 billion annually.

The bulk of nondefense emergency spending was provided for disaster relief in most years of the 1991-1999 period, but nearly every major domestic department or independent agency received emergency funds at some time (see Table 2). Nondefense emergency spending ranged from $1.5 billion (1991) to $13.8 billion (1999) and averaged about $6.9 billion annually.
 

Certain accounts have received emergency funding in more than one year and others in virtually every year since the emergency designation has been available. The Departments of Defense (DoD) and Housing and Urban Development (HUD) have received emergency funding in every year since 1991. HUD's emergency funding has usually been provided for contributions for assisted housing or community development block grants. DoD's funds have been used for a variety of accounts and purposes. The Federal Emergency Management Agency has received emergency funding in every year since 1992 for disaster relief and disaster assistance loans.

Emergency funding for the Small Business Administration's disaster loans has been provided in most years since 1992 in both regular and supplemental appropriations, depending on the timing of the disaster to which the loans respond. Similarly, the Department of Agriculture's Commodity Credit Corporation has received emergency funding since 1992. That funding was most often provided in supplemental appropriations but was provided in regular appropriations in 1995.

A few accounts tend to receive most emergency funding in regular appropriations. The Low Income Home Energy Assistance Program (LIHEAP) has been designated an emergency in regular appropriations in every year since 1992. Regular appropriations for the fire-fighting operations of the Forest Service and the Department of the Interior have been designated as emergency requirements each year since 1993 (except for 1998).

For 1999, emergency funding has been provided for a variety of purposes, in some cases for the first time. Defense emergency funding is included as in previous years for overseas contingency operations (mostly for the ongoing operations in Bosnia) and for costs resulting from natural disasters at U.S. military installations (hurricanes in the United States and flooding in South Korea). However, emergency funding was provided for the first time, for example, for ballistic missile defense enhancements ($1 billion), "to address funding shortfalls and other requirements" in the defense health program ($200 million), and for defense counter-drug and drug interdiction activities ($42 million).

Nondefense emergency funding for 1999 is provided for crop losses as in years past, but for the first time the emergency designation has been used for price supports for farmers. Funds are provided "to partially compensate producers for loss of markets in 1998 due to circumstances beyond their control, such as regional economic dislocation, unilateral trade sanctions, and failure of the government to pursue trade opportunities aggressively." Other nondefense emergency funding is provided for the first time for the year 2000 computer conversion, security enhancements at various federal properties, and additional counter-drug and drug interdiction activities.

Until fiscal year 1999, most emergency funding was in the form of supplemental appropriations. Those supplementals generally appeared in stand-alone bills (in which the entire public law was designated emergency spending) or in larger supplemental bills (in which a title or a handful of accounts carried the emergency designation). In contrast, emergency spending for 1999 was scored as regular appropriations because it was packaged along with eight regular appropriation bills in the 1999 Omnibus Appropriations Act.

In most years, the bulk of emergency appropriations were designated by both the President and the Congress at the time they were enacted. However, most emergency amounts for 1996 and 1999, and about half for 1998, were initially enacted as contingent emergencies. As of November 19, 1998, about $10.4 billion in contingent emergencies still awaited designation by the President. Of that amount, $7.6 billion was provided in the 1999 Omnibus Appropriations Act.

The use of the emergency designation can also be measured by the number of times the emergency designation has been invoked (see Table 1) and by identifying the appropriations subcommittees that have jurisdiction over the emergency spending, the agencies that receive the emergency funding, and whether the President or the Congress originated the spending (see Table 2).⁽⁵⁾ (Tables in the appendix display that information separately for supplementals and for regular appropriations.)

Offsetting Emergency and Nonemergency Supplementals

Since the enactment of the Deficit Control Act of 1985, the Congress has sought to constrain the effects of supplemental spending by including offsetting rescissions in supplemental appropriation acts (see Figure 2, which shows supplemental appropriations, net of rescissions, as a share of total discretionary budget authority). Under the BEA, the Congress has enacted rescissions to fully offset nonemergency supplemental spending in most years. In some years since 1994, the Congress has offset emergency supplemental spending as well.
 

FIGURE 2. SUPPLEMENTAL APPROPRIATIONS NET OF RESCISSIONS, 1985-1998

SOURCE: Congressional Budget Office.

NOTE: Through the end of the 105th Congress (October 21, 1998), emergency spending was provided as part of regular appropriations for the year.

Through the end of the 105th Congress (October 21, 1998), all emergency spending for 1999 had been included with regular appropriations in the 1999 Omnibus Appropriations Act. That act rescinds $2.7 billion for 1999, but those rescissions were not explicitly designated to offset emergency spending. Since Figure 2 shows offsets to supplemental spending only, it does not include emergency spending for 1999.

Trends in Domestic Emergency Spending

Two earlier CBO analyses summarized data on supplemental spending for the 1970s and 1980s.⁽⁶⁾ Figure 3 shows those data as well as data on domestic emergency spending for 1991 through 1999 (excluding defense and international spending). The data, however, are not strictly comparable. Before 1991, they reflect only supplemental spending related to natural disasters (for example, the Northeast blizzard, floods, and drought in 1978; the eruption of Mount St. Helens in 1980; and Hurricane Hugo and the Loma Prieta earthquake in 1989 and 1990). In contrast, the data for 1991 and thereafter include both regular and supplemental appropriations as well as spending for emergency purposes other than natural disasters (for example, the Los Angeles riots, the Oklahoma City bombing, and additional funds for continuing drug interdiction efforts).
 

FIGURE 3. SUPPLEMENTAL SPENDING FOR NATURAL DISASTERS (1976-1990) AND DOMESTIC EMERGENCY SPENDING (1991-1999)

SOURCE: Congressional Budget Office.

NOTES: The figure shows the disasters and emergencies that largely accounted for the spikes in spending.

Pre-1991 data are from Congressional Budget Office, Supplemental Appropriations in the 1970s (July 1981), and Supplemental Appropriations in the 1980s (February 1990). Post-1990 spending includes both regular and supplemental appropriations. All data exclude defense and international discretionary spending.

Since 1991, domestic emergency spending has been high relative to the disaster-related spending of the 1970s and 1980s. There are several possible explanations. Some analysts would attribute the higher levels to the combined effects of tight discretionary caps under the BEA and the emergency safety valve (see the next section). Indeed, domestic emergency spending in 1999 is near the peak for such spending in the post-BEA era.

However, others believe that much of the domestic emergency spending since 1991 reflects an unusual incidence of losses rather than changes in federal budget practices. For example, emergency funding for 1992 responded to the riots in Los Angeles, flooding in downtown Chicago, and Hurricanes Andrew and Iniki. The large spike in 1994 was spending driven by the Northridge earthquake. Emergency funding for 1995 and 1996 went for antiterrorism initiatives in the wake of the Oklahoma City bombing. Funding was appropriated in 1997 to respond to massive flooding in North and South Dakota and in 1998 to ice storms in the Northeast and tornadoes, floods, and El Nino-related weather conditions elsewhere. At least part of the $6 billion in 1999 agricultural emergency funds was intended to help farmers who experienced crop losses as a result of droughts and floods, but some of the funding may also be used to protect farmers from the effects of falling commodity prices.

Effects of the Discretionary Caps on Emergency Spending

The discretionary spending limits of the past decade have generally capped total discretionary spending below inflation-adjusted levels. Many policymakers feel that those limits can encourage relatively higher levels of emergency spending since such spending is effectively exempt from the caps. The record levels of emergency spending for 1999 have only heightened their concern.

Some policymakers feel that emergency spending has increased when the underlying discretionary caps have been changed by new budget agreements. The BEA originally specified caps on discretionary spending through 1995. Those caps have been extended twice, in 1993 and 1997. The Omnibus Budget Reconciliation Act of 1993, which carried out the 1993 budget agreement, defined new caps for 1996, 1997, and 1998 but made no changes to the caps for 1994 and 1995. The 1997 Balanced Budget Act, which added new caps for 1999 through 2002, changed the caps governing 1998 to make them more closely match the levels of discretionary budget authority and outlays agreed to in the 1997 budget agreement. That change meant a $6.9 billion decrease in the 1998 cap on budget authority and a $6.9 billion increase in the 1998 cap on outlays.⁽⁷⁾ Emergency spending for 1998 was in the lower range for the post-BEA years, excluding spending for Operation Desert Storm (see Figure 1).

Other policymakers are more concerned that emergency spending tends to be higher in years when the discretionary spending limits are more restrictive. In general, the discretionary caps under the three multiyear budget agreements of the 1990s become more restrictive in the later years of those agreements. But evidence of higher emergency spending in those later years is mixed. For example, in the last two years covered by the 1990 budget agreement, total emergency spending (excluding Operation Desert Storm) fluctuated from relatively high levels (1994) to levels in the middle range for the post-BEA period (1995). In fiscal year 1997, the next-to-the-last year of the 1993 budget agreement, emergency spending was in the middle to higher range for the BEA era.

Whether the unprecedented level of emergency spending for 1999 is a harbinger or an aberration is unclear. The 1999 amounts were enacted in the second year of the 1997 balanced budget agreement. They were included as regular appropriations for the year, not subsequently in separate supplemental appropriations as is the norm.

Nonetheless, if the emergency safety valve is being used to increase funding for nonemergency activities, the pressure to do so will only increase. The discretionary caps for the last three years of the agreement--fiscal years 2000 through 2002--tighten significantly; they generally cap total discretionary appropriations below the level enacted in 1999.⁽⁸⁾ Under that scenario, projected budget surpluses for the next few years, when combined with restrictive caps, would increase the pressure on policymakers to rely on the BEA's emergency safety valve to provide levels of total discretionary spending exceeding the caps.

However, the BEA establishes no definition of emergency, provides no means for evaluating whether emergency designations are justified, and does not cap emergency spending. It requires only that both the President and the Congress designate emergency amounts. Furthermore, determining the amount that the various factors may have contributed to higher emergency spending in some years is difficult. Whether certain emergency spending--in 1999 or earlier years--was provided for nonemergency purposes is largely a matter of subjective judgment.
 

OPTIONS FOR CHANGING THE BUDGETARY TREATMENT OF EMERGENCY SPENDING

Opinions about the current budgetary treatment of emergency spending are mixed. Critics of the current system feel that objective criteria, based partly on experience, should be established to guide and constrain decisions about emergency spending. Critics are also concerned that the emergency designation effectively constitutes a large loophole that weakens overall budgetary discipline. Some also claim that excessive emergency aid discourages the purchase of adequate insurance and weakens incentives to take actions that would mitigate the cost of natural disasters.

Defenders of the current procedure fear that many of the proposals to improve the planning and control of emergency spending could lead to the underfunding of legitimate emergency needs and would shift too much authority to the President. One advantage of the current system, they assert, is that the Congress takes an active role in deciding the merits of various requests for emergency spending. They also maintain that the emergency safety valve is essential for holding multiyear budget agreements intact. The discretionary spending limits have become increasingly restrictive in recent years, which they claim makes an emergency safety valve even more necessary. They note that in some years since 1994 the Congress has offset emergency supplemental appropriations with spending cuts.

Several options have been considered in recent years for changing the budgetary treatment of emergency spending. Those options would eliminate the BEA emergency exemption and effectively require emergency spending to be offset, retain the emergency exemption but require a supermajority vote of the Congress to approve emergency spending, establish specific criteria for designating spending as an emergency requirement, or create a reserve fund for emergency spending.

Eliminate the Emergency Exemption

One approach for controlling emergency costs is simply to eliminate the exemption for designated emergencies under the BEA. Emergency spending would be counted under the discretionary spending limits or included in the PAYGO estimates and calculations, as appropriate. Emergency appropriations would have to compete with other funding priorities in regular appropriation acts; emergency spending provided in supplemental appropriation acts during the fiscal year would have to be offset with spending cuts.

Policymakers may choose to offset emergency spending costs but retain the emergency exemption. As mentioned earlier, since 1994 the Congress has routinely included offsets for certain emergency supplementals without being required to do so by budget law. Eliminating the emergency exemption and requiring offsets by law could make the process less flexible. Without the exemption, a major domestic disaster in a heavily populated area or a massive overseas troop deployment, for example, could require several billion dollars in additional spending that would force large offsets in other programs.

Require a Supermajority Vote for Emergencies

Another approach would be to retain the BEA emergency exemption but require a supermajority vote, perhaps a three-fifths majority, to approve any emergency spending. The Senate, for example, currently employs a series of three-fifths voting requirements to waive various enforcement provisions under the Congressional Budget Act. One advantage of this approach is that it would necessitate a change only in Congressional procedure, not in the BEA, and thus would not need the President's concurrence.

Advocates of a supermajority vote maintain that it would create a more rigorous test for policymakers while retaining the current emergency safety valve. A true emergency, they claim, would easily garner the necessary three-fifths majority. Opponents counter that such a requirement would give too much power to a determined minority in a single House of the Congress. They are also concerned that it would put the onus for emergency spending discipline on the Congress alone, whereas the President would be able to recommend any amount of emergency spending without penalty.

Establish Criteria for Emergency Spending

The BEA does not set forth any criteria for policymakers to use in designating spending as an emergency requirement. Some people advocate establishing specific standards that would have to be met before spending could be declared an emergency requirement under the act. Such criteria, they say, would make decisions about emergencies less subjective, would guard against abuses of the emergency designation, and would strengthen budgetary discipline and control.

In 1991, the Office of Management and Budget drafted the following five criteria to use in deciding whether funds should be designated as an emergency requirement:

• Necessary expenditure--an essential or vital expenditure, not one that is merely useful or beneficial;
  
  
• Sudden--quickly coming into being, not building up over time;
  
  
• Urgent--a pressing and compelling need requiring immediate action;
  
  
• Unforeseen--not predictable or anticipated as a coming need; and
  
  
• Not permanent--the need is temporary.

In theory, those or other criteria would shrink the range of spending provisions that would qualify for the emergency requirement and make it more difficult for such provisions to be enacted. Those criteria might also provide additional information for policymakers to consider and might give opponents objective grounds on which to contest emergency spending. However, it is unlikely that any set of criteria would completely remove the subjective element from policymakers' decisions about emergency spending. Furthermore, it is possible that such criteria would be ineffective and would simply appear as a few standard paragraphs in each emergency supplemental appropriation act.

Establish a Reserve Fund for Emergencies

The federal government could set up an emergency reserve fund similar to those used by some state governments.⁽⁹⁾ In general, a reserve fund would consolidate most or all emergency spending into a single fund or budget account and would provide budgetary resources in advance of emergency needs. However, establishing a reserve fund for emergencies, by itself, would not resolve the question of whether emergency spending should be exempt from BEA requirements. If policymakers establish a reserve fund and eliminate the BEA emergency exemption, they will have to find ways to accommodate or pay for emergency needs under the discretionary spending limits and the PAYGO requirement.

The government, however, does not need to establish a reserve fund to budget and provide sufficient funds for emergencies. No procedural hurdle prevents policymakers from increasing regular appropriations for anticipated emergency needs, as long as their actions are consistent with the discretionary spending limits. However, one advantage of using a separate reserve fund is that it might highlight overall emergency needs more effectively and enable policymakers to draw a more direct connection between emergency spending and any offsets used to pay for that spending. An emergency reserve fund could also be used to encourage efforts to avoid or mitigate disasters as well as to highlight potential alternatives to federal action, such as state or local initiatives or private insurance.

Proposals to establish a reserve fund have come from both the President and the Congress. The President proposed a contingent emergency reserve account for domestic emergencies in his 1998 budget. His 1998 appropriation request for the fund ($5.8 billion) was based on the average annual discretionary cap adjustments made for such emergencies since 1991. However, the President proposed that the requested amount, and any additional amounts that might be needed later, should continue to be exempt from the BEA's discretionary caps.

Several bills establishing similar procedures for an emergency reserve fund were introduced (but not acted on) in the 105th Congress. H.R. 457 would establish a budget reserve account funded through annual appropriation acts and controlled under Congressional budget resolution procedures. It would also eliminate the BEA exemption for emergency spending and require that such spending be budgeted with other appropriations under the discretionary spending limits. H.R. 1372 would require the budget resolution to include a separate functional category for spending on natural disasters. It would also require that the level of disaster assistance spending recommended in the resolution be allocated to committees and enforced by points of order. If disaster assistance spending exceeded the amounts allocated, the President would be empowered to rescind spending in an amount equal to the excess.

As part of its overall recommendations on reforming the budget process, the House Budget Committee Task Force on Budget Process from the 105^th Congress recommended establishing a reserve fund for emergencies. The task force's recommendations are embodied in H.R. 4837, which was introduced near the end of the Congress. Under that bill, budget emergency would be defined in law as an unanticipated situation that requires federal spending to mitigate, prevent, or respond to "loss of life, property, or a threat to national security." The budget committees, in consultation with the relevant legislative committees, would be directed to establish and print guidelines for applying the statutory definition. The President's budget and the budget resolution would be required to include emergency spending levels as a separate spending category (divided into discretionary and mandatory amounts). The spending levels would be based on a five-year rolling average. When the House or Senate considered legislation that contained emergency spending, the budget committee Chairman would be required to certify whether those amounts were for an emergency as defined by law. Only then would he or she be authorized to adjust levels and allocations in the budget resolution to accommodate the emergency amounts. Emergency spending bills that exceeded those levels would be referred to the budget committee to decide whether they should be exempted from the caps or the PAYGO requirement.

The concept of a reserve fund raises a number of important questions and issues, including how much to provide to the fund, how it should be controlled, and how or whether its use should be enforced under current budgetary disciplines. For example, given the wide fluctuation in emergency spending over the years, what is the appropriate amount for the fund, and for what period should amounts remain available? Who should control the release of those funds and under what conditions? If the President is given the authority to control the funds, should that authority be circumscribed, and if so, how? Finally, is any budgetary discipline or fail-safe mechanism needed to ensure that the fund is not abused or used for unintended purposes?

Funding. A reserve fund is intended to eliminate the need for supplemental appropriations, essentially by budgeting for those supplemental amounts in advance. However, it is impossible to know how much emergency spending will be needed in any given year. One solution would be to make a permanent, indefinite appropriation to the fund from which most or all emergency needs could be drawn as needed. That arrangement would eliminate the uncertainty of funding, but it could create a large and relatively uncontrollable source of funds for federal agencies that would not be subject to annual review by the Congress.

Another approach would be to appropriate an annual amount to the fund based on a historical average spent during a prior period of five or 10 years. That approach is already used for fire-fighting programs of the U.S. Forest Service and the Department of the Interior and has been used as the basis for appropriation requests for the FEMA Disaster Relief Fund. However, in calculating the average amounts spent on disaster relief, the Administration has excluded certain very costly disasters, such as Hurricane Hugo and the Loma Prieta and Northridge earthquakes of the early to mid-1990s. Even though the large amounts spent on those disasters seemed to be funding aberrations at the time, excluding them nonetheless understates the calculation of average historical costs and raises questions about how such costs should be determined.

If some historical average was used as a benchmark to fund emergency reserves, any balances remaining in the fund at the end of the fiscal year could be carried over for qualified spending in later years--the approach used for fire-fighting programs. In the case of a single reserve fund for all emergencies, carryover balances could become quite large if spending needs in any single year were much less than appropriated. Thus, a critical issue is how and under what conditions the amounts in the reserve fund, including any carryover balances, should be controlled.

Control. One option for controlling balances in the reserve fund would be to allow them to lapse at the end of each fiscal year. That approach would tighten control of the funds, but it could defeat the purpose of providing advance emergency funding and could lead to excessive spending just before the funds lapse at the end of the fiscal year.

Another option would be to allow balances in the fund to carry over but to give the President control over how those balances should be obligated. Under the President's proposal for an emergency reserve fund in his 1998 budget, amounts in the fund would be available for obligation 15 days after the President notified the Congress of his intent to use them. The President would be given flexibility to use the funds for disaster assistance programs as needed. That approach could be more efficient than the current practice of having the Congress control supplementals, but it would shift budgetary power away from the Congress. It also presumes that the President is less likely than the Congress to spend excessively or wastefully for emergency needs.

A different approach would be to center control of a reserve fund in the Congress through the budget process. For example, a reserve amount could be recommended in the budget resolution and allocated to the appropriations committees or other committees of jurisdiction. That approach was used in H.R. 4837 (105th Congress). Alternatively, the Congress could use a reserve fund procedure that delays the decision about how much to set aside for an emergency reserve. For example, an additional allocation for emergency spending to the appropriate committees of jurisdiction could be made under the budget resolution only when legislation offsetting that spending was considered or adopted by the Congress. Postponing the decision about how much should be appropriated to the reserve fund could, however, make it more difficult to find sufficient offsets if emergency spending needs were unexpectedly large.

Enforcement. Another issue involves whether separate budget enforcement disciplines should be created to constrain emergency spending. Regardless of whether policymakers decide to retain the BEA emergency exemption, they could choose to establish procedural controls to ensure that emergency spending is held to budgeted amounts.

One such device, which could be used in conjunction with an emergency reserve fund, would be to create a "fire wall" for emergency discretionary spending that is enforced by separate caps on discretionary spending. Separate caps are now in place for defense and nondefense spending, certain discretionary spending for highways and mass transit, and the Violent Crime Reduction Trust Fund (the separate defense and nondefense caps are merged after 1999, and the separate violent crime caps expire after 2000). A separate cap for emergencies could be carved out of total discretionary spending as well, with the other caps reduced commensurately by the total amount reserved for emergencies. To ensure that amounts reserved under the emergency caps would be appropriated only for emergency purposes, policymakers could identify in law or in committee report language (as they do now for other capped spending categories) the specific budget accounts that may be funded under those caps. Emergency spending appropriated in excess of the caps would then trigger across-the-board reductions either in all or in a select group of discretionary spending programs.

However, wide fluctuations in total emergency spending from year to year might make caps unrealistic and inflexible. Policymakers have revised the discretionary spending limits only twice since 1991 (in 1993 and 1997). More frequent revisions would probably be necessary if caps were set on emergency spending. In some years, emergency spending needs might exceed the caps by large amounts, and the resulting sequestration or spending offsets could force large reductions in other programs. In other years, when emergency spending fell well below capped levels, policymakers might be tempted to fund other programs with amounts reserved under the emergency spending caps.
 

CONCLUSION

It is impossible to predict the catastrophic events or unexpected developments that lead to emergency spending. Yet experience informs us that they will indeed occur and on a fairly regular basis. Generally, lawmakers have dealt with unpredictable events in ad hoc fashion, providing the bulk of emergency funds as they are needed. But some lawmakers wonder if that approach is best. They are concerned that demands for emergency funding in the midst or aftermath of a crisis, when public pressure for funding is strong, may lead to unnecessary or wasteful spending. They feel that changes in the budgetary treatment of emergencies may help control costs. Others are skeptical and fear that proposed budgetary changes could make it more difficult for a consensus of lawmakers to provide timely assistance.

Budgeting for emergency spending is inherently difficult and uncertain. Emergency funds are provided for a wide variety of purposes, are administered by many agencies, and are generally unpredictable for any particular fiscal year. Policymakers have therefore acknowledged the need for a budgetary safety valve for emergencies as a part of recent budget enforcement disciplines. Many are concerned, however, that the safety valve has served as an excuse to avoid programmatic planning in many instances and has provided a budgetary loophole for excessive spending. The liberal use of the emergency exception in the 1999 Omnibus Appropriations Act has highlighted those concerns.

A reserve fund for emergencies might promote better planning in the budget process. Whether it could also help control costs is unclear. Fashioning a budgetary mechanism that completely eliminates the need for emergency spending will be difficult. Making changes in emergency programs--for example, to incorporate measures to mitigate costs--might be one approach to easing the budgeting task, but uncertainty is likely to remain a central element of that process.

Ultimately, the options for changing the budgetary treatment of emergency spending are limited by the current budget enforcement process. As long as the discretionary spending caps and PAYGO requirement remain in effect, policymakers face a fundamental choice: to exempt emergency spending from those budget enforcement disciplines or to find ways within those requirements to offset unexpected emergency costs.
 

APPENDIX

This appendix includes tables with additional information on emergency spending under the Budget Enforcement Act since 1991. Tables A-1 and A-2 show emergency spending in supplemental and regular appropriations. Tables A-3 and A-4 show the number of designated emergencies in supplemental and regular appropriations. Table A-5 lists all the supplemental and regular appropriation acts that include emergency spending.
 

1. Congressional Budget Office, Supplemental Appropriations in the 1980s (February 1990), p. 5.

2. No separate procedures were established to enforce those caps. Instead, they were enforced through the legislative process and the President's veto power.

3. William G. Dauster, "Budget Emergencies," Journal of Legislation, vol. 18, no. 2 (1992), pp. 251-258.

4. Only two emergencies have been declared under the PAYGO process. The Emergency Unemployment Compensation Amendments of 1993 included an emergency designation for extended unemployment benefits provided in that law (sec. 6 of P.L. 103-6). CBO estimated that emergency mandatory spending in the act would increase outlays by about $5.7 billion in fiscal years 1993 and 1994. The Internal Revenue Service Restructuring and Reform Act of 1998 included an emergency designation for a provision that waives interest penalties on underpayments of income taxes filed by taxpayers in a Presidentially declared disaster area (sec. 3309(c) of P.L. 105-206). CBO estimated that the provision would lower revenues by about $130 million in fiscal years 1998 through 2003.

5. The information in Table 2 and the appendix on whether the President or the Congress originated the spending should be used with caution. Such a distinction is quite difficult to draw, a point highlighted in the recent negotiations on emergencies in the 1999 Omnibus Appropriations Act. The information reflects available documentation, which consists of any written request from the President that was transmitted before the emergency spending was enacted. If such documentation was not available, CBO assumed that the request originated in the Congress.

6. Congressional Budget Office, Supplemental Appropriations in the 1970s (July 1981), and Supplemental Appropriations in the 1980s (February 1990).

7. Budget of the United States Government, Fiscal Year 1999: Analytical Perspectives, Budget Enforcement Act Preview Report, Table 14-1, p. 260.

8. See Congressional Budget Office, Final Sequestration Report for Fiscal Year 1999 (October 30, 1998).

9. For background on emergency spending practices in the states, see the statement of Theresa A. Gullo, Chief, State and Local Government Cost Estimates Unit, Congressional Budget Office, before the Task Force on Budget Process, House Committee on the Budget, June 23, 1998.