What's Wrong with the Federal Budget Process

Report Budget and Spending

What's Wrong with the Federal Budget Process

January 25, 2005 37 min read Download Report
Brian Riedl
Brian Riedl
Senior Fellow, Manhattan Institute

The current federal budget process is failing to meet its most basic obligations. The process is supposed to provide an orderly roadmap for determining the nation's annual spending and revenue priorities, but instead it stifles debate, prevents cooperation, and frequently breaks down. Created in 1974, the current process has been subject to 30 years of abuse and loopholes from lawmakers hoping to exploit its structural flaws. The budget process finally collapsed during the fiscal year (FY) 2003 and 2004 budget debates, during which:

  • The annual budgets were not completed until four
    months into the fiscal year that they were designed to fund;
  • The House of Representatives, Senate, and White House could not even agree on a basic budgetary framework for fiscal year 2005;
  • President George W. Bush neared completion of his 2004 budget submission-and most agencies laid the groundwork for their 2005 budget requests-without knowing their 2003 budget levels; and
  • Discretionary spending caps that were created to limit spending were allowed to expire without any replacement.

With so much still undecided, the safest prediction is that, as in previous years, the eventual winners of this year's budget battle will be those who can read the legal fine print and twist the available loopholes and gimmicks in their favor. This is no way to write a budget.

Budgets are about setting priorities. Nations often have infinite desires and only finite resources to fulfill them. A rational budget progress should help lawmakers set priorities and separate vital needs from unaffordable luxuries. Without spending limits and priority setting, lawmakers lack the tools to allocate resources efficiently, and the overall result is wasteful spending, high taxes, and sluggish economic growth.

Why Reform Is Difficult

Despite the popularity of fixing the budget process, the last eight years have seen every major reform proposal defeated in the House or Senate. One reason for this failure is that nearly all proposals inherently alter the relative budget-writing influence of some members, committees, or legislative bodies, and those losing any of their power fight tirelessly to defeat the proposal. Another cause of failure is the lack of urgency for budget discipline controls during a recent period of soaring government revenues and budget surpluses. This justification for complacency has crumbled: The war on terrorism and homeland security costs are putting new pressures on spending. Ensuring a disciplined budget process that best allocates the federal government's resources is now a matter of national security and economic health.

The past 30 years have provided valuable lessons regarding which process reforms succeed, which fail, and why. Because many of the past reforms succeeded until lawmakers exploited their structural faults, reforming the budget process does not require reinventing the wheel, but only repairing it. The 1974 Congressional Budget and Impoundment Control Act required that all annual appropriations bills be coordinated under the umbrella of a broad budget resolution that set general budgetary priorities. The Budget Enforcement Act of 1990 (and the prior Gramm Rudman Hollings Act) gave budgeters multi-year targets within which to work. These reforms brought rational planning and coordination to the budget process, and their subsequent weakening has made it, once again, disjointed and chaotic. Reforms strengthening the concepts of this prior legislation will revitalize the budget process. The most important reforms would:

  • Place a strong yet realistic and flexible cap on each year's total spending
  • Resolve major differences earlier in the process by bringing the President into the annual congressional budget resolutions
  • Put all programs (including mandatory ones) on the table during the annual appropriations debate; and
  • Strengthen budget enforcement by requiring a two-thirds vote to bypass any of the budget restraints.

Problems with the Current Budget Process

The current budget process has four major problems.

Problem #1: The President and Congress have little incentive to seek agreement. In addition, Congress habitually violates its own internal agreements. Presidents and Congresses seldom agree on even the broadest budgetary blueprint; therefore, negotiating an agreement on the funding of the hundreds of programs that require annual appropriations is a daunting and time-consuming task. Given the scale and scope of this endeavor, it is senseless that the President and Congress are not required to begin negotiating their differences until the very end of the budget process, when the appropriations bills reach the President's desk. Without having previously agreed on an overall budgetary blueprint, the chances of Congress and the President negotiating a timely agreement on the details of each program are slim. Accordingly, each autumn the President and Congress are faced with a choice between two undesirable options: (1) hastily negotiating their differences regarding hundreds of programs (affecting billions of dollars) without the benefit of any binding framework; or (2) investing the time needed for careful negotiations but exceeding the budget completion deadlines, thereby leaving many agencies at risk of running out of funding. The current budget process not only discourages early agreements between Congress and the President, it discourages Congress from enforcing its own internal agreements. Until their 2002 expiration, multi-year discretionary spending caps were supposed to constrain congressional budget resolutions, which in turn constrain appropriators. Another spending safeguard was the pay-as-you-go (PAYGO) model, which was created to ensure that neither the expansion of entitlements nor tax legislation would increase the budget deficit. However, both the discretionary caps and PAYGO rules could be overridden by a simple majority vote in the House and a three-fifths vote in the Senate.[1]Constraints that can be discarded this easily are not authentic constraints. A better budget process would bring the relevant players to the bargaining table early in the year to set a broad-yet binding-framework that would guide the decisions regarding specific program appropriations throughout the course of the year. The process should also be workable and realistic and should not impede the daily functions of the federal government if negotiations stall.

Problem #2: The budget process fails to provide a clearinghouse for all spending. The federal budget is supposed to provide an opportunity for Congress and the President to step back and decide how much the federal government should tax and spend during the following year. In reality, only the one-third of spending that is classified as discretionary is subject to the appropriations process every year. Almost all other programs (classified as mandatory) are left without regular oversight to grow uncontrollably from year to year. Thus, the budget process denies the nation's policymakers an opportunity to set annual spending and tax priorities with all programs on the table. Furthermore, the mandatory programs that are "off the table" represent the largest long-term threat to the nation's fiscal health. In 2008, the first Baby Boomers will begin collecting Social Security and Medicare benefits, with costs expected to increase enough to raise federal spending by 5 percent of gross domestic product (GDP) by 2030 (the current equivalent of $5,200 per household annually) and 13 percent of GDP by 2050 (the current equivalent of $13,500 per household annually).[2]This would result in either substantial tax increases or elimination of most other federal programs. A rational budget process cannot simply ignore this issue by taking these problems off the table. Mandatory programs are not the only ones excluded from the budget process. Although natural disasters and other emergencies occur nearly every year, they are not anticipated in the budget process. Congress regularly allocates all available budgetary resources to non-emergency expenses. When the inevitable $5 billion to $15 billion disaster relief tab reaches Congress each year, it has no choice but to exceed the spending levels of the original cap. Although some catastrophic emergencies may be too large to budget for, there is no reason why Congress cannot set aside funds within each budget for non-catastrophic emergency expenses.

Problem #3: The budget process is designed with a bias toward higher spending and taxes. Public choice theory recognizes that how democracies make decisions has a substantial effect on what is decided. Multi-year constraints, such as PAYGO and discretionary spending caps, represent an attempt by policymakers with a long-term view to constrain the decisions of annual budgeters who are focused only on the short term. However, these multi-year constraints fail to settle the question of whether the budget process should be used to limit spending (as discretionary caps suggest) or to slow the growth of the budget deficit, regardless of government size (as PAYGO suggests). This confusion created odd situations whereby even policies that would achieve both goals of reducing spending and reducing the budget deficit (such as a discretionary spending cut accompanied by a smaller tax cut) have not been allowed. Furthermore, PAYGO did not successfully blunt the pro-spending bias of annual budget writers because it focused only on the effects of new policies and ignored current policies-because it was rarely enforced. Public choice theorists also note the pro-spending bias caused by the decentralization of spending committees. Although a single appropriations committee in the House and the Senate annually approves all discretionary spending, nearly a dozen different committees in each body of Congress write mandatory spending programs. The lack of coordination between these committees creates a "tragedy of the commons," whereby each committee is responsible only for the funding of its own pet programs with no obligation to trade off their costs with the costs of other committees' programs.[3] Accordingly, each committee over-prioritizes and consequently over-funds its own programs. A single committee reviewing all legislation would solve this bias by taking on the responsibility to make the difficult trade-offs.[4]

Odd Accounting Rules.
Two congressional accounting rules also tilt the playing field in favor of higher taxes and spending. When deciding how to adjust a program's annual funding, Congress is required to start from "current policy" baselines that have factored in new costs from inflation, increased benefits, and new enrollees. If the current policy baseline for a program mandates a 10 percent increase, then even an allocation that is 9 percent above the previous year's levels is considered a "cut." Policymakers wanting to avoid the erroneous impression of "cutting" funding for popular programs must vote for larger spending increases than they may deem necessary. Accounting rules also inflate the appeal of higher taxes. When calculating the effect of a proposed tax cut, Congress assumes that higher taxes will not depress any sector of the economy (thereby reducing the new revenue created) and that lower taxes will not provide any economic stimulus that could recover some of the lost revenue. This bias always makes higher taxes appear more appealing than lower taxes from a budgetary and economic standpoint.

Problem #4: The budget process is complex and subject to abuse by the few who can master its intricacies. The current budget process is too complex for voters, journalists, and even many policymakers to understand. As a result, budgetary decisions are often dominated by those who can take advantage of its intricacies and bend the complex budget rules in their favor. Rather than highlighting such matters as the impact of policies and budgetary priorities, attention is focused on such terms as advance appropriations, obligation and payment delays, and points of order. A reformed budget process should be simple, easy to implement, and less prone to loopholes and gimmicks.



A Proposal for Reform

The order of the following recommendations matches the timeline of the federal budget process. The proposal begins with the outside statutory constraints placed on annual budgets before they are written. It then moves through an annual budget process, with the budget resolution followed by the appropriations process and presidential approval. The proposal concludes with a section on general reforms that would affect all portions of the budget process.

Reforming Outside Constraints on Annual Budgets

Although Congress and the President write a new budget each year, policymakers have long acted to constrain the choices that budgeters face. These outside constraints can be constitutional, such as the constitutional requirement of Article I, Section 8, that government spending be preceded by congressional appropriations and the long-standing proposal for a constitutional amendment to require a balanced budget. Statutory laws also provide external budget constraints, such as those requiring Congress to write an annual budget resolution or to adhere to long-term spending caps and PAYGO limits.

Permanent, rigid, external constraints on annual budgets are considered necessary because:

(1)They bring predictability and order to the annual budget process; (2)multi-year planning allows policymakers to debate long-term budget goals; and (3) in a given budget debate, they can require policymakers to make the necessary hard choices. Critics of outside restraints argue that it is unwise to restrict the options of future budgeters, especially given the difficulty of anticipating future budgetary and economic situations.[5] A simple, easily implemented budgetary framework that would allow the flexibility needed to meet the critics' objections is described below.

Recommendation #1: Congress should place annual caps on total spending rather than bringing back discretionary spending caps and PAYGO.

The current round of multi-year budget restrictions began in 1985, when the Gramm-Rudman-Hollings Act (GRH) enforced annual budget deficit targets by allowing the President to sequester excess spending to meet the targeted budget deficit. In 1990, policymakers concluded that GRH budget deficit targets were too strict to be realistic and yet too weak to reduce the budget deficit. They replaced GRH with a system of multi-year caps on discretionary spending and PAYGO rules requiring that any new entitlement or tax legislation be deficit neutral. If legislation violating caps or PAYGO was enacted, the Office of Management and Budget (OMB) would be required to enact across-the-board sequestrations on all non-exempt programs until the cost of the legislation was balanced. Although caps and PAYGO expired in late 2002, Congress appears inclined to bring them back.[6]

The Promises and Pitfalls of Discretionary Caps. The relatively strict discretionary caps of the early 1990s mandated that discretionary spending grow at approximately 2 percent per year¾much slower than the 5 percent annual rate of the 1980s.[7] Congress generally obeyed these caps and reduced the growth rate of discretionary spending (with the unavoidable exception of the Persian Gulf War). Caps showed promise despite their flaws. For instance:

  • Discretionary caps succeeded best in the short run. They succeeded because they were simple and understandable. They failed when they ceased being realistic. Throughout the 1990s, caps were typically set for five-year periods. This proved to be too long a time horizon to anticipate the needs of an evolving economic and budgetary scenario. Thus, the five-year series of caps created in 1990 was rewritten by 1993. The new five-year caps set in 1993 were significantly altered in 1996 and then replaced in 1997 by a new series of five-year caps that were also effectively abandoned by 2000. Clearly, a series of spending caps cannot be expected to have a shelf life beyond three years before the changing economic and budgetary picture renders it outdated and unrealistic.
  • Discretionary caps lacked real enforcement. Caps become unrealistic when Congress decides that the spending reductions they require are too politically dangerous. Such situations expose a second weakness of past spending caps-how easily they can be ignored. These multi-year constraints are supposed to provide a check on the whim of the short-term congressional majority. Therefore, one would expect that they would be enforced with additional hurdles that Congress must clear before bypassing them. Yet, rather than raising the bar from a majority vote to a supermajority to bypass caps, the rules now only require Congress to designate additional spending as "emergency" spending before passing it with a simple majority.
  • The emergency designation was intended for use only when unforeseen disasters and catastrophes required additional spending. However, the loophole invited abuse by making it easy for legislators to bypass spending caps without a supermajority vote.In the 2000 budget, Congress designated $44 billion of mostly ordinary spending as "emergency" to bypass the outdated cap levels.[8] Congress actually used the emergency designation properly in the 2001 and 2002 budgets by limiting it to the $20 billion appropriated each year for legitimate 9/11-related emergency spending. Legislators did not actually obey the caps those years: They simply used a different loophole-raising the spending caps by $96 billion in the 2001 budget and then by $135 billion in the 2002 budget-to accommodate all extra spending demands.[9]

The solution is not to abolish caps, but to recognize their proper use and realistic limitations. Caps were intended to constrain congressional spending and to force legislators to make the difficult decisions required for responsible budgeting. There are limits, however, on legislators' willingness to accept politically unpopular decisions forced upon them by the budget process. Furthermore, caps will fail to force Congress to make even slightly inconvenient choices if they can be overridden by a simple majority vote. These problems are fixable.

Shortcomings of PAYGO. In contrast to discretionary caps, PAYGO was conceptually flawed from the start. PAYGO mandated that any new tax or mandatory spending legislation not increase the budget deficit. This policy required that the costs of any tax reductions or entitlement expansions must be balanced either by raising taxes in other areas or reducing other entitlements. If this counterbalance was not accomplished, at the end of the year the Office of Management and Budget would sequester other spending in the budget to a level that equaled the costs incurred by the new legislation. PAYGO had four major shortcomings:

  • PAYGO was not enforced. No meaningful sequestration ever took place in 12 years of PAYGO, in part because Congress was careful to avoid PAYGO-violating legislation. Like caps, however, PAYGO was ignored when it was deemed unrealistic. Between FY 1998 and FY 2001, Congress enacted legislation violating PAYGO by a total of $46 billion and then attached provisions to later legislation preventing OMB from enforcing the law with a sequestration. Congress then voted to block the sequestration of its $130 billion violation for FY 2002 and its $127 billion violation for FY 2003.[10] Thus, PAYGO served as only a slight deterrent to new tax and entitlement legislation.
  • PAYGO prevented trade-offs. Lax enforcement was not the only factor in PAYGO's failure. Another was the inconsistent treatment of dis-cretionary funding and mandatory funding. Discretionary funds were subject to caps on spending, but mandatory spending was allowed to increase as long as taxes were raised correspondingly. The separate rules applied to the different categories of expenditures prevented any trade-off between discretionary spending and mandatory spending or taxes. For example, policymakers could not shift $10 billion in discretionary expenditures to mandatory programs, because this would violate PAYGO. Conversely, shifting $10 billion from mandatory to discretionary spending would violate the caps. Even though balancing a $10 billion tax cut with a $20 billion reduction in discretionary spending would both reduce spending and the budget deficit, such a move would not be permissible because it would violate PAYGO.
  • PAYGO ignored current mandatory spending. Not only did PAYGO actually prevent policies that would have tamed the budget deficit and curtailed excessive spending, it was also designed to allow mandatory programs to grow unrestrained. PAYGO prevented new legislation from increasing the budget deficit. Yet most of the growth in mandatory spending that caused the budget deficits and excessive federal spending over the past 25 years arose from the natural, automatic growth of current entitlement programs such as Social Security, Medicare, and Medicaid. PAYGO centered the mandatory spending debate around the dangers of new legislation, but treated the entire $1.5 trillion spent on current mandatory programs-many of which are expensive, outdated, and wasteful-simply as a given.[11] The preference of PAYGO's designers to place current mandatory spending outside of the law's reach is a major reason why mandatory spending (excluding net interest) actually grew faster in the 12 years with PAYGO than in the previous 12 years before the law (See Chart 1).[12]


  • PAYGO raised taxes.The same PAYGO rules that allowed entitlement spending to grow unrestrained also made tax increases automatic. The booming 1990s increased the incomes of millions of workers, thereby pushing them into higher marginal income tax brackets. PAYGO's goal of deficit neutrality implies that legislation merely reversing these unlegislated tax increases should be allowed. Yet PAYGO acknowledged only tax and entitlement changes resulting from legislation. Therefore, legislated tax cuts could not be used to correct unlegislated tax increases, even if it is within the deficit-neutral goal of restoring revenues to their original level. Despite the best of intentions, PAYGO's legacy has been a bias toward high taxes, an inability to address runaway entitlement growth, and counterproductive administrative rules discouraging trade-offs between mandatory and discretionary spending.

Total Spending Caps: A Key to Workable Spending Limits. The most simple and effective policy to ensure responsible budgeting would replace the current collection of discretionary spending caps and PAYGO rules with one simple cap on total spending for each year. This cap could take either of two forms:

  1. TABOR caps would be determined by a formula. Modeled after Colorado's Taxpayer Bill of Rights (TABOR) law, a TABOR cap would limit annual spending increases to the inflation rate plus population growth.[13]
  2. OmniCaps would be set manually by lawmakers. These caps would function like an authorization. Every three years, the budget committees would write legislation authorizing a maximum level of spending for each of the next three years. Each year's OmniCap would consist of just two numbers-a budget authority total and an outlay total.[14] The authorization would become a binding law upon passage by the full House and Senate and the President's signature.[15]

Each year's spending cap would be enforced through the budget resolution, whose total spending target would be limited to the current year's cap.[16] A resolution exceeding the year's cap would require a two-thirds vote by Congress instead of the regular majority vote. Each annual budget resolution would then go to the President's desk to be signed or vetoed. Once the budget resolution's total spending target is in place, the budget process would continue as usual: The budget resolution would set spending breakdowns; authorizers and appropriators would perform their duties (with the same two-thirds vote needed to exceed any of the 302(a) allocations or 302(b) sub-allocations); and the President would sign or veto the appropriations bills.

Total spending caps have several significant advantages over the current system:

  • Simplicity. Each year's cap would contain just one amount for budget authority and one amount for outlays¾no gimmicks, loopholes, or sophisticated and overly complex reforms.
  • Flexibility. Congress would be able to write each year's budget without the kind of micromanagement that has prohibited trade-offs between discretionary and mandatory spending. Spending could be allocated across programs in any way that lawmakers see fit so long as total spending remains below the current year's cap.
  • Competition. Unlike the PAYGO system, total spending caps would put current as well as proposed mandatory spending on the negotiations table. This provision ensures that all spending requests will compete with one another and that no item will receive a free pass.
  • Reasonable enforcement. The requirement of a two-thirds vote to override these caps would be easily achievable in the event of a catastrophe, such as the 9/11 attacks, but would make it too difficult to exceed spending limits when Members of Congress simply wish to avoid making tough choices. Requiring a presidential signature for the annual budget resolution serves as an additional hurdle to unjustified overspending.

Critics contend that any cap encompassing mandatory programs will lead to drastic cuts in vital entitlements. Under Omnicaps, that would occur only if lawmakers consciously choose to set low OmniCap levels. Under a TABOR cap, it would occur only if lawmakers refuse to rein in low-priority programs to make room for priority entitlement programs.

Reforming the Annual Budget Resolution

Prior to 1974, annual spending and revenue totals were not centrally coordinated or even planned by Congress. Each spending and tax bill was individually written, debated, and passed, and Congress waited until the end of the entire process to tally up the total. The 1974 Congressional Budget and Impoundment Control Act[17] created annual budget resolutions through which Congress would commit to total spending and revenue targets before jumping into annual appropriations and reconciliations.[18] Over time, however, the scope and enforcement of budget resolutions have weakened to the point at which annual appropriations are beginning to resemble the chaotic pre-1974 process. The following proposals will strengthen budget resolutions and their role in setting annual targets.

Recommendation #2: The total spending cap should limit the annual budget resolution's spending levels, and the annual budget resolution should limit annual appropriations.

For multi-year spending restrictions to be relevant, they must be enforced all the way down to individual spending bills. To that end, total spending caps should serve as a ceiling on annual budget resolution spending levels, which, in turn, should serve as a ceiling on annual appropriations and mandatory expenditures.[19] This means that bypassing the total spending cap, the annual budget resolution's spending target, or the appropriations 302(b) sub-allocations should require a two-thirds voteby both the House and the Senate. This would ensure that the limits are bypassed only in times of real emergency.

Recommendation #3: The President should be included in formulating the budget resolution.

If two sides are expected to work out a detailed agreement on a complex subject within a nine-month period, it makes little sense to separate them for the first eight months. It makes even less sense for one side to spend so much time working out the smallest details of its offer without having first forged the basic structure of an agreement with the other side. Yet this is how the federal budget is currently written. The President begins the process in February by presenting his proposed budget as an opening offer. Congress then spends up to eight months preparing its counteroffer in the form of 13 detailed annual appropriations bills. At that point, with the deadline for completion quickly approaching, the President's options are limited to making a sign-or-veto decision on each appropriations bill. At this stage, negotiations prove to be extremely difficult because the two sides must haggle over the details of hundreds of programs in the absence of a basic framework of aggregate spending targets. Inevitably, this process results in rushed compromises that are completed well past the fiscal year deadline.

An increasingly popular solution would be to move from a concurrent budget resolution (which does not involve the President) to a joint budget resolution (which would require the President's signature to become law). By working out the differences early in the process and enacting a binding law, the President and Congress would have a chance to settle broad issues regarding the appropriate level of total spending and revenue. This would limit the appropriations debate to the composition of federal spending, and any disagreements in this area would be far easier to work out with the spending totals already in place.[20]

There will be times when Congress and the President cannot agree on spending and revenue levels. The budget process could be moved forward in those instances by allowing Congress to initiate appropriations bills under an unsigned budget resolution. Requiring a two-thirds vote to pass each appropriations bill in the absence of a joint budget resolution will provide adequate pressure for Congress and the White House to settle their differences.

ome critics contend that giving the President a role in the budget resolution unfairly transfers power from Congress to the President. However, Congress's budget cannot be enacted until the President approves it.[21] This reform would simply move up the inevitable negotiations and provide an opportunity to settle contentious issues earlier rather than later.

Recommendation #4: Spending should be broken down by committee (and appropriations subcommittee) instead of by function.

Current budget resolutions follow the President's lead by breaking down spending into functional classifications such as international affairs, transportation, and income security.

A more relevant breakdown is the amount that each committee and subcommittee will be allowed to spend. While committee breakdowns are provided in the budget report, they are not the focus of deliberations and they exclude the subcommittee breakdowns that are vital for the appropriations committee. Rather than distracting Members with the functional breakdown, Congress should write a budget resolution that focuses solely on the committee and subcommittee breakdown. A simple one-page document providing total spending limits and a committee-by-committee (as well as appropriations subcommittee) breakdown would allow Congress to see and debate the budget resolution in the format most relevant to the subsequent appropriations process and would provide a roadmap detailing each committee's expectations.[22]

Although Congress should structure the budget resolution around the committees, this does not mean the current committee setup cannot be improved. Specifically, many of the 13 appropriations subcommittees are grouped by issues totally unrelated to each other (such as veterans, housing, and NASA for one subcommittee, commerce, justice, and state for another). If lawmakers grouped the appropriations subcommittees by their spending functions (such as health, income security, and transportation), they could more rationally and systematically determine and implement their budget priorities.

Reforming the Annual Appropriations Process

After completing the budget resolution, the respective budget committees divide up the budgetary resources by committee. Committees that authorize mandatory and tax legislation rarely are required to act because their programs will continue operating under current formulas (although some budgetary resolutions contain reconciliation instructions requiring a committee to rewrite its programs to achieve a specified level of budgetary savings). The most important allocations are to the House and Senate appropriations committees, which receive all discretionary spending allocations and then provide a further subdivision to each of their respective 13 subcommittees. Only expenditures for these programs are scrutinized annually.

Recommendation #5:Mandatory spending should be brought into the annual appropriations process.

In the absence of reconciliation instructions (which are becomingly increasing rare), only the one-third of federal spending classified as discretionary must move through the annual appropriations process. The other two-thirds, classified as mandatory spending, is simply deemed uncontrollable and set on autopilot to grow without any restraints or oversight. In the absence of long-term planning and accountability, mandatory spending since 1962 has grown 11 times faster than discretionary spending.[23] Even supporters of increased mandatory spending should agree that the federal budget cannot provide a meaningful national debate on spending and tax priorities if two-thirds of the federal government's programs are immediately taken off the table. Labeling programs untouchable and providing them with blank checks insulates them from the trade-offs that must be made when competing priorities are weighed.

Moving all spending into the annual appropriations process would allow Congress to make the trade-offs necessary to promote the most efficient and effective use of resources. All programs would be sent to the appropriations committee to receive a specific dollar appropriation for the upcoming year and would be forbidden from spending beyond that amount. Although it is likely that most mandatory programs would escape major reforms, Congress would be granted the tools to rein in programs that spend more than is deemed necessary.

Answering the Critics. Critics point out the difficulty of appropriating the right amount to mandatory programs. Indeed, most mandatory programs have historically been excluded from the appropriations process because their program beneficiaries are legally "entitled" to their benefit, leaving the cost of the program to be determined by the number of enrollees. In reality, most entitlement programs have predictable enrollment changes from year to year, making it relatively easy for Congress to anticipate the funding levels needed to maintain current benefits.

The Food Stamp program proves that not all entitlements require a blank check. Currently, Congress and the President annually decide how much the nation should spend on this program for the following year, and then appropriate that amount. Recipients remain entitled to their benefits, and if the annual appropriation appears insufficient to fund all beneficiaries at current benefit levels, then those levels may be slightly adjusted to keep spending in line with the appropriation. All other mandatory spending could be appropriated in this manner, except for the payment of net interests on government bonds, which the federal government is obligated to pay to its investors. Political realities may keep Social Security and Medicare exempt from the appropriations process as well.[24]

Committees that authorize mandatory programs may oppose sharing their spending power with the appropriations committee. In a sense, appropriations committees would merely take over the reconciliation duties that currently allow budget committees to order reductions in mandatory spending. The real difference between the two systems is that reconciliation allows authorizing committees to find their own spending reductions, while this proposal empowers the appropriations committee to write the actual reductions. Still, it is unlikely that appropriators would make significant changes to mandatory spending formulas during the appropriations process without asking the authorizing committee to revisit the law themselves.

Another concern might be that the appropriations committees would be overwhelmed by the duties of appropriating all programs. Biennial budgeting presents one way to ease that pressure. Bringing the President into the budget resolution process in the spring should also reduce the number of issues that would have to be worked out with the White House in the autumn, thus allowing for smoother and quicker negotiations. Furthermore, it is unlikely that appropriators would significantly adjust many mandatory programs in a given year.

However, if lawmakers insist on leaving mandatory spending outside the appropriations process, TABOR or OmniCaps could still apply. Lawmakers could continue to budget for those programs in the budget resolution. If unanticipated mandatory expenses during the fiscal year threaten to push total spending above the overall cap, lawmakers could simply go back and find savings either in those programs or elsewhere in budget. This could be accomplished manually or automatically by writing a "trigger" into the budget resolution specifying where to streamline spending in such an event.

Recommendation #6:Congress should create an emergency reserve fund.

Most families understand the wisdom of setting aside money for unexpected emergencies. Likewise, states maintain "rainy-day funds" for disaster relief and other unforeseen needs. The federal government, however, does not set aside any funding for emergencies. More than just an oversight, the failure to budget for emergencies provides Congress with a backdoor means of exceeding spending caps. After allocating all budgetary resources to regular spending programs within the limitations of caps, Congress can later claim that "unforeseen" emergency needs required it to exceed the original spending caps.

The current budget process makes this easy. As stated earlier, Congress can bypass discretionary spending caps to fund emergencies with only a majority vote. Exceeding the budget resolution's spending ceilings requires only a majority vote in the House and a three-fifths vote in the Senate.[25] These low hurdles for excess spending provide Congress with little incentive to fund emergencies (or even non-emergencies) underneath discretionary spending caps or the annual budget resolution.

Congress should budget sufficiently for regular emergencies, while assuring that the necessary escape hatch for catastrophic emergencies is not abused. Such reforms would promote better planning and coordination within the constraints of the budget caps and annual budget resolution.

How It Would Work. Given that non-war emergency spending throughout the 1990s averaged $9 billion per year, Congress could require that the annual budget resolution allocate a minimum of 0.5 percent of the total annual budget (approximately $10 billion in 2004) as an emergency reserve fund. Allowing unused emergency budget authority to roll over to the following year would ensure that the fund could remain solvent even if annual emergency spending fluctuates between $5 billion and $15 billion annually, so long as average annual emergency funding does not exceed $10 billion per year. In the event that emergency spending beyond the capacity of the fund is required, these expenditures should be offset by spending cuts in other areas. Requiring that all emergency spending must come from a separate appropriations bill and meet a stipulated definition of "emergency" will help to prevent abuses of the fund.

It is possible that a catastrophic event could drain the emergency fund and require expenditures that are too large to be offset by other spending reductions. In such a situation, additional emergency spending that bypasses the statutory total spending caps and the budget resolution should require the same two-thirds vote in Congress that would be required for non-emergency spending.

No "Emergency" Loopholes to Abuse. A two-thirds vote is a high hurdle, but the chain of budgetary enforcement is only as strong as its weakest link. If the vote required to exceed the budget resolution for emergency spending is any lower than the budget resolution's non-emergency hurdle (two-thirds), or the hurdle for exceeding multi-year total spending caps (two-thirds), Congress would likely once again abuse the emergency loophole to bypass spending limits (See Table 1). In the event that a massive catastrophe exhausts both the emergency fund and the additional amount that could reasonably be offset elsewhere in the budget, Congress should have no problem producing the necessary two-thirds vote to appropriate extra funding.


Recommendation #7:
Congress should eliminate advance appropriations.

In addition to designating all new spending as "emergency," Congress often uses advance appropriations to exceed spending limits. Advance appropriations allow legislators to provide budgetary authority for funding that would not become available until a year or more after the current fiscal year. For example, if legislators writing the 2005 federal budget run out of room under the current spending cap and budget resolution, they can allocate additional expenditures for the 2006 fiscal year. Advance appropriations are similar to spending caps in that they are a form of multi-year planning. However, there is a significant difference between the two mechanisms: Whereas spending targets provide a broad framework within which future Congresses must create their budgets, advance appropriations bind future Congresses to designated funding allocations for specific programs. In doing so, they place a tight and unnecessary straitjacket on future budgeters.

Advance appropriations averaged under $3 billion per year in the 1990s, but expanded to nearly $24 billion by 2004 as legislators used them to exceed spending caps (See Chart 2).[26] Lawmakers often defend the $15 billion in annual advance appropriations for education programs by asserting that schools operating on a different fiscal year than the federal government need early assurances of full funding. Yet even the Department of Education admits there is no programmatic justification for advance appropriations.[27] The 2005 federal budget (which runs from October 2004 through September 2005) funds the 2005-2006 school year (which runs from July 2005 through June 2006). Thus, even without advance appropriations, nearly all education spending becomes available well before the school year begins. Furthermore, the predictability of annual federal education spending allows local school districts to plan their budgets in advance.


Likewise, $9 billion in annual advance appropriations is not justified in areas such as housing vouchers and Head Start. Congress could eliminate advance appropriations by requiring that all budget authority be credited to the current or following year's spending totals-regardless of when the funds are actually released.

Recommendation #8: Service on the appropriations and budget committees should be limited.

Most of these proposals to bring order to the budget process would increase the power of the budget and appropriations committees. The House and Senate budget committees would write the OmniCaps (although not a TABOR cap), as well as the more strongly enforced annual budget resolutions. The House and Senate appropriations committees would annually appropriate funds for mandatory as well as discretionary programs. The powerful influence of these committees on the federal budget, which affects every issue, should be balanced by a steady rotation of membership. Limiting Members to four years on each committee out of every 12 years they serve in Congress would ensure regular turnover. These restrictions need not be unnecessarily strict: A Member could serve four years on each committee, even at the same time. Prior committee service in the other body of Congress should not count against a Member's total.

In addition to diffusing the committees' increased power, committee term limits would address the bias that typically emerges over time on these committees. For example, Capitol Hill observers have long noted that increased tenure on appropriations committees often creates a bias toward excessive spending by obscuring a big-picture perspective. Conversely, the big-picture focus of budget committees often results in their members' having a more fiscally conservative orientation than their colleagues outside these committees. Increasing the turnover on these committees will help them to more accurately represent the views of Congress as a whole.

Recommendation #9: Congress should prevent government shutdowns with automatic continuing resolutions.

Even the best efforts to simplify the budget process and foster cooperation cannot guarantee that a budget will be completed by the annual October 1 deadline. Failure to complete the budget by this deadline risks paralysis in the federal government because it leaves many agencies unable to spend money. Congress has missed the October 1 deadline in 26 of the past 27 years, resulting in a series of continuing resolutions (CRs) to fund the federal government in the interim. Contentious debate regarding the composition of the CRs creates regular uncertainty among both providers and recipients of federal services.[28] This insecurity could be assuaged by a law that guaranteed continued funding at the prior year's rate to agencies without a budget for the new fiscal year. Some in Congress may oppose using the prior year's budget: Many fiscal liberals would support automatic increases, such as cost of living adjustments, and many fiscal conservatives would support automatic reductions for one-time expenses or even an entirely different budgetary formula. Simplicity, however, makes the prior year's funding level the most practical option. Budgeters who dislike that formula would have an added incentive to forge a new budget agreement more quickly.

Reforming the Role of the President

The most important presidential power in the budget process is the authority to sign the appropriations into law or to veto them. Given the difficulty of overriding a presidential veto (which requires a two-thirds majority in both the House and the Senate), Congress must write a budget that the President will be willing to sign.

Because Congress is not obligated to incorporate the President's budget proposal into the budget resolution or any appropriation, the President is on the sidelines throughout much of the budget debate. Although the White House can attempt to influence Congress by asserting its viewpoint, the President does not directly influence the final budget until the end of the process when the appropriations bills are submitted for the President to sign or veto. The President could be brought into the process earlier by converting the concurrent budget resolution into a joint budget resolution that would be signed by the President. A number of other popular proposals have been made to strengthen the President's role at the end of the budget process through impoundment, enhanced rescission, and a constitutional line-item veto.

Recommendation #10: Presidential rescission should be strengthened.

The increasing usage of omnibus spending bills is forcing lawmakers to vote on increasingly large spending bills. Consequently, the same bills that fund priorities such as homeland security, health, and education also fund $50 million for an indoor rainforest in Iowa and $250,000 for the Rock and Roll Hall of Fame in Ohio. The only course of action available to lawmakers who refuse to fund wasteful spending is to vote against the entire bill.

Rescission bills provide an innovative way to cut the wasteful spending out of legislation. After the legislation is enacted, the President can request a bill to de-fund wasteful programs and grants before the funds are spent. Lawmakers frustrated by having to vote for the wasteful components of an otherwise good bill are given an opportunity to go back and remove the waste.

Congress, however, often balks at rescission requests. Rather than go on record as supporting wasteful pork-barrel projects and obsolete programs, Congress blocks rescission requests by refusing to vote on them.

A positive reform would guarantee that all presidential rescission requests receive "expedited consideration" by proceeding directly to the House and Senate floor for a vote. Rather than continuing to avoid rescission requests, lawmakers would be forced to go on record as supporting or opposing these wasteful expenditures. The likely result would be less wasteful spending.

Enhanced rescission is a weaker presidential tool than a line-item veto (which could be overridden only by a two-thirds vote of Congress) or impoundment (whereby a President could rescind spending without any input from Congress). Nonetheless, it is a feasible way to help the President check wasteful spending without removing any significant congressional powers.

Accounting Changes Across the Budget Process

Several accounting requirements currently provide Congress with a biased perspective of the costs and benefits of tax and spending proposals. They should be replaced by mo.re accurate accounting methods that better represent the real-world impact of budget proposals.

Recommendation #11: The annual budget should include liabilities and unfunded obligations.

The current budget process is prepared on a cash basis, in which spending is based on when cash is expected to change hands. There is no budgetary consideration of the actual financial commitments resulting from policy changes enacted by Congress because these will be paid well into the future. As a result, entitlement programs such as Social Security and Medicare have huge unfunded obligations attached to them. Because they are financed on this pay-as-you-go basis, lawmakers are not forced to deal with this problem in a serious way. The private sector, on the other hand, has a regimented process for acknowledging and tackling these types of obligations. The underlying principles of private sector financial reporting should be applied to the federal budget process to present a clearer picture of the nation's fiscal health.

The proposals below are presented as a menu of options-by no means exhaustive-to add to the current debate about dealing with long-term entitlement obligations.

Limiting Growth in Federal Liabilities and Obligations. Adoption of a budget, or limit, for liabilities and obligations as part of the budget process would allow lawmakers to make policy changes, but require them to live within the limits established by a liability budget. Any new spending policy would have to be evaluated on the liability or obligation it creates over a long-term basis. This would provide policymakers with a measure of full disclosure and transparency that they now lack when making financial decisions. To live within the "budget," Congress would pay for liability increases by cutting other obligations.

Adding a "Payment Plan" for Liabilities and Obligations to the Annual Budget. Adding a payment measure for liabilities and obligations to the annual budget would force Congress to acknowledge the cost of federal obligations and begin to balance near-term spending patterns with long-term realities by reducing spending growth. This would work by including an "annual payment" of federal obligations in the budget resolution. In times of budget deficits this "annual payment" would require policymakers to budget more responsibly by reserving resources that would otherwise be used for spending, although it would not directly reduce total liabilities and obligations. In any year with a budget surplus, resources would be reserved to pay down liabilities such as federal pension liabilities or debt held by the public. This measure would work best in tandem with strong, enforceable spending limits to prevent lawmakers from merely increasing deficits.[29]

Recommendation #12: Congress should terminate baseline budgeting.

The 1974 Congressional Budget and Impoundment Control Act requires Congress to use "current policy" baselines as the starting point for consideration of new spending.[30] For many entitlement programs, projected spending calculated from these baselines is automatically adjusted for inflation, new enrollees, and increased benefits. Under this system, accounting standards have defined annual increases as large as 7 percent as spending "freezes" and spending increases as high as 6 percent as "reductions." Although a 6 percent spending increase may not be enough to fully fund a program under certain standards, simple arithmetic shows that it isnot a reduction. Baseline budgeting discourages Members of Congress from voting for smaller spending increases in popular programs, given that such a vote could be characterized as a reduction in funding. Honest and responsible budgeting should not include such misleading adjustments.

Recommendation #13: Tax proposals should be scored using reality-based models.

Accurate budget planning requires using the best tax projections available. Congress, however, currently requires that its scorekeepers assume that tax rates do not affect the economy. They assume that tax reductions will not stimulate economic growth or recover a portion of the lost tax revenue and that tax increases will not harm the economy, thereby diminishing the anticipated revenue gain. These "static" scoring models, which are rejected by most economists, consistently overstate the positive budgetary effects of high taxes. Using any of the existing "reality-based" scoring models would include the economic effect of tax proposals and lead to more accurate tax projections.


The chaos of the of the 2003 and 2004 federal budget debates and the expiration of discretionary spending caps and PAYGO rules provide Congress with both a reason and an opportunity to step back and comprehensively examine the current budget process. The past 30 years have shown that a budget process does not last longer than 10 to 12 years before mounting abuses necessitate major reforms. The last major reforms, enacted in 1990, hung on by a thread before expiring in 2002.

Although war and economic revitalization continue to dominate the agenda, leaving little time for comprehensive budget process reform, Congress should resist the temptation to reflexively vote to continue the current process. This process-which has engendered shortsighted, haphazard decision making by those who can manipulate the process-will not efficiently allocate the nation's resources to its highest, crucial priorities: defense, anti-terrorism, economic recovery, and long-term entitlement reform.

Brian M. Riedl is Grover M. Hermann Fellow in Federal Budgetary Affairs in the Thomas A. Roe Institute for Economic Policy Studies at The Heritage Foundation.

[1] A three-fifths vote in the Senate is required only if a Senator raises a point of order against legislation for violating these rules. Otherwise, the Senate can override these rules with a simple majority.

[2] Heritage Foundation calculations based on Congressional Budget Office, "Long-Term Budget Outlook," December 2003, at www.cbo.gov/showdoc.cfm?index=4916&sequence=0 (July 9, 2004).

[3] The budget committees are intended to address this "tragedy of the commons" by writing annual budget resolutions that trade-off all spending. Budget resolution writers may attach reconciliation instructions requiring that authorizing commit­tees reduce their mandatory program budgets. However, budget committees rarely issue such reconciliation instructions. They instead focus most of their attention on determining how much discretionary spending to allocate to the appropria­tions committees.

[4] Professor John Cogan of Stanford University examined both decentralized and centralized budgeting processes between 1799 and 1989. During the 97 years of centralized budgeting, the average annual budget deficit was only 0.15 percent of the GDP. In the 94 years of decentralized budgeting, the average annual budget deficit was 2.49 percent of GDP. This data supports the position that the current decentralization of the federal budget process leads to increased government spend­ing. John Cogan, "The Federal Budget," in David R. Henderson, ed., The Concise Encyclopedia of Economics, Web ed., 2002, at www.econlib.org/library/Enc/FederalBudget.html (July 9, 2004).

[5] Proposed constitutional amendments that would require a balanced budget or a supermajority to raise taxes are good ways to provide significant multi-year discipline. Such reforms are beyond the scope of this paper, which focuses on what Con­gress and the President can accomplish in the short term without amending the U.S. Constitution.

[6] The Senate has temporarily retained the rule allowing a Senator to offer a point of order against legislation violating PAYGO, which requires a three-fifths vote to override. Violating PAYGO would not induce sequestration.

[7] Calculated from Office of Management and Budget, Budget of the United States Government, Fiscal Year 2003: Historical Tables, Table 8.2, page 124, and Office of Management and Budget, Budget of the United States Government, Fiscal Year 2003: Analytical Perspectives, p. 287.

8] Calculated from the Office of Management and Budget, Analytical Perspectives, p. 287.

[9] Ibid. Because spending caps are set by statute, altering them requires enacting a new law. Congress would write the new cap into the final appropriations bill during each year.

[10] Office of Management and Budget, Analytical Perspectives, p. 291, and Bud Newman, "Senate Passes Pay-Go Bill Ordering OMB to Avoid Sequester of Entitlement Programs," BNA Daily Report for Executives, November 19, 2002, p. G5.

[11] Rudolph Penner, a former director of the Congressional Budget Office, defends PAYGO's policy of focusing only on legis­lated changes in mandatory spending. He states that the annual changes in mandatory spending are driven mostly by eco­nomic factors, and when a weak economy automatically triggers increased mandatory spending, it becomes politically infeasible to trim these programs enough to remain under the required level. Although the experience of the Gramm-Rud­man-Hollings Act fits within this viewpoint, the nation cannot afford to continue allowing mandatory spending programs to grow at their current rate without any real oversight or limits. If past limits on mandatory spending proved unrealisti­cally tight, as Penner suggests, policymakers could write looser caps. Shortening the time horizon of caps to three years will also keep them more realistic than past caps. Loose and realistic caps are better than no caps at all. Rudolph G. Penner, "Repairing the Congressional Budget Process," Urban Institute, May 2002, at www.urban.org/url.cfm?ID=410522 (January 10, 2005).

[12] Calculated from Office of Management and Budget, Historical Tables. Annual mandatory spending (excluding net interest) grew by 40 percent between 1980 and 1991 (before PAYGO) and by 51 percent between 1991 and 2002 (during PAYGO).

[13] For a more complete explanation of the proposed federal Taxpayers' Bill of Rights, see Brian M. Riedl, "Restrain Runaway Spending with a Federal Taxpayers' Bill of Rights," Heritage Foundation Backgrounder No. 1793, August 27, 2004, at www.heritage.org/Research/Budget/bg1793.cfm.

[14] Legislators would be free to include non-binding guidelines for breakdowns such as mandatory and discretionary spend­ing.

[15] One potential danger lies in Congress's having to delay the beginning of the annual budget once every three years because it must first complete the OmniCap reauthorization (i.e., delaying the 2006 budget until Congress completes the Omni­Caps for 2006, 2007, and 2008). Enacting these caps the year before Congress begins writing the first of its three budgets (2005 in this example) would leave time to write those budgets. However, writing them too early would require Congress and the President to project spending levels several years ahead, which would risk rendering the later OmniCaps unrealis­tic.

[16] Multi-year revenue targets do not accompany multi-year spending caps because targets are nearly impossible to enforce. The federal government can firmly control spending because it dispenses the funds itself. Revenues, on the other hand, come from taxpayers and are dependent on the individual working, spending, and savings patterns of 290 million Ameri­cans. Governments can set tax laws and aim to keep revenues within a certain range but they cannot enforce specific tar­gets. Nor does a feasible remedy exist if revenues begin to diverge from their annual target, because enacting and implementing new tax policies would take too long to affect the current fiscal year. Therefore, tax revenue goals fit best in the annual budget resolution, where they serve as non-binding guidelines.

[17] Biennial budgeting would provide Congress with more time to better plan and debate each budget and to conduct more stringent program oversight. However, it is generally excluded from this proposal because vehement congressional opposi­tion has made biennial budgeting a "third rail" of budget process reform, immediately killing any proposal that includes it. Biennial budgeting is not required for this paper's proposals to succeed.

[18] Public Law 93-344.

[19] In addition to requiring a two-thirds supermajority to exceed the 302(a) cap on total discretionary spending, lawmakers could apply that supermajority requirement to legislation exceeding the 302(b) subcommittee allocations.

[20] An innovative idea promoted by Representatives Chris Cox (R-CA) and Paul Ryan (R-WI) would allow Members of Con­gress to reduce spending limits toward the end of the budget process. Currently, Members who reduce a program's funding while debating an appropriations bill on the floor see those savings quickly spent on different programs. The "lock box" proposal would allow the amendment's authors to retain the savings by reducing the bill's spending limit by the amount of the proposed spending reduction. This seems fair: The current budget process allows Congress to easily raise spending limits at the end of the budget process, so it should also be able to lower them. This reform would not fit into this paper's proposal because it violates the central idea of settling spending totals early in the process. Easily adjusting them upward or downward later in the process would once again create a situation in which the President, the House, and the Senate end up with different spending totals that must be negotiated at the end of the process. Rather than achieving symmetry by giving both sides the ability to effortlessly alter spending limits, neither side should be allowed to do so. Of course, Con­gress could always decide to spend less than the OmniCap or TABOR cap.

[21] Presidential vetoes are occasionally overridden by a two-thirds vote in the House and Senate, which happens to be the same vote required to exceed budget restraints in this proposal.

[22] The following section's proposal to move all spending into the annual appropriations process would mean all spending (except net interest) would be allocated to the appropriations committees. Therefore, the relevant breakdown would be by appropriations subcommittees.

[23] Brian Riedl, Rea Hederman, and Ethan Baker, Federal Revenue and Spending: A Book of Charts (Washington, D.C., The Heri­tage Foundation, 2002), Chart S-13.

[24] Congress should make adding to this list of exempt programs as difficult as possible. The hurdles should include requiring a strong supermajority vote and presidential signature, as well as banning "off-budget" program labels.

[25] Again, a three-fifths vote in the Senate is required only if a Senator raises a point of order against legislation for violating these rules. Otherwise, a simple majority vote could override these rules in the Senate.

[26] Committee on the Budget, U.S. Senate, "Informed Budgeteer," No. 21, 107th Cong., 2nd Sess., July 9, 2002, and data pro­vided by the Senate Budget Committee.

[27] Ibid.

[28] This uncertainty was heightened after the inability of Congress and the President to agree on a CR led to a government shutdown in the winter of 1995-1996.

[29] See Alison Fraser, "Time for the Federal Budget Process to Include Unfunded Entitlement Obligations," Heritage Founda­tion Backgrounder, forthcoming.

[30] Joint Economic Committee, U.S. Congress, "Budget Process Reform," May 1997, p. 6.




Brian Riedl
Brian Riedl

Senior Fellow, Manhattan Institute