Hearings of the House Committee on Rules
H.R. 853, The Comprehensive Budget Process Reform Act of 1999
Executive Director, Center for Budget and Policy Priorities
I appreciate the invitation to testify before the committee. I am Robert Greenstein, director of the Center on Budget and Policy Priorities. The Center is a non-profit, non-partisan policy institute here in Washington that specializes in fiscal policy and in policies and programs affecting low- and moderate-income families and individuals. The Center is funded primarily from foundation grants and receives no funding from any part of the federal government. By way of background, I also would note that in 1994, I had the privilege of serving with distinguished Members such as Rep. Goss on the Kerrey-Danforth Bipartisan Commission on Entitlement and Tax Reform.
Some of the provisions of H.R. 853, the Comprehensive Budget Process Reform Act, would make useful improvements in the federal budget process. I would take special note of the legislation's proposed changes in the budgetary treatment of federal insurance programs.
Unfortunately, the legislation also contains provisions that would have quite undesirable effects. Various provisions of H.R. 853 could make the budget process less efficient than it is today and lead to long delays in action on appropriations bills. H.R. 853 also could lead to larger reductions in discretionary programs than the reductions already envisioned under the 1997 Balanced Budget Act. In addition, it could lead to automatic cuts in Medicare and other programs during economic slowdowns and weaken budget mechanisms that are preserving budget surpluses which may turn out to be needed for Social Security reform. Overall, I believe the legislation would damage the budget process significantly and that its drawbacks substantially outweigh its advantages. I would recommend against enacting this legislation.
- The bill would be likely to squeeze discretionary programs inordinately. It would likely have this effect because of the combined effect of a number of features of the legislation. H.R. 853 would effectively allow the caps on discretionary spending to be reduced to pay for tax cuts and entitlement expansions. It also would lower the discretionary caps when either the House or Senate passed an amendment to an appropriations bill reducing funding for a discretionary program.
-
The bill would likely lead to delays in consideration of appropriations bills because it would repeal the provision of current law allowing appropriations bills to be brought to the House floor after May 15 if a budget resolution has not yet been approved; it would bar floor action on appropriations bills until work on the budget resolution has been completed. Moreover, the bill would lengthen the time it takes to finish work on a budget resolution because it would convert the resolution into a joint resolution that requires a Presidential signature.
-
Under H.R. 853, if projected surpluses are used to pay for tax cuts and the surpluses subsequently fail to materialize as forecast -- which could easily happen if the economy performs less well than forecast or tax cuts turn out to be more expensive than was assumed at the time they were enacted -- cuts in Medicare, student loans, farm price supports, and various other entitlements would be triggered through the sequestration process.
-
The bill would alter the "pay-as-you-go" rules, allowing projected surpluses in the non-Social Security budget to be used to finance tax cuts and entitlement increases before Social Security reform is approved and before it is known whether a portion of these funds are needed to fashion Social Security or Medicare solvency legislation that can secure majority support in both houses.
I will elaborate on these points below. Before doing so, I would like to caution that we should be very careful about making large changes in the budget rules established under the Budget Enforcement Act of 1990. While it is sometimes said that the budget process is broken, most budget experts I know think otherwise and believe the Budget Enforcement Act of 1990 has been remarkably effective. The regimen it established of discretionary caps and pay-as-you-go requirements has been instrumental in helping us get from multi-hundred billion dollar deficits as far as the eye can see to budget surpluses.
When the BEA was enacted, some predicted the discretionary caps would routinely be busted by large amounts and the pay-as-you-go requirements would not last long. Both predictions proved mistaken. That the emergency designation of the law was stretched last fall and again in the current Kosovo supplemental reflects the fact that the 1997 budget agreement set unrealistically austere caps, not that the process as a whole has broken down. Moreover, changes to tighten the procedures for handling emergencies and making emergency designations can be instituted without the more sweeping changes that H.R. 853 would make.
Let me turn to what I regard as the principal shortcomings of the legislation.
Impact on Discretionary Programs
Discretionary programs constitute a declining share of the budget. At $575 billion in fiscal year 1999, discretionary spending accounts for 34 percent of the budget and seven percent of the economy (i.e., of the Gross Domestic Product). The Congressional Budget Office projects that if discretionary spending stays within the caps through 2002 and grows with inflation thereafter, discretionary spending will decline to 29 percent of the budget and five percent of GDP by 2009. By contrast, ten years ago in 1989, discretionary spending constituted 43 percent of the budget and nine percent of GDP.
Various provisions of H.R. 853 would directly or indirectly place additional downward pressure on funding for discretionary programs. The bill contains a "lock-box" provision that would cause reductions in the discretionary caps. After the House and Senate had completed floor action on any appropriations bill but before conference on the bill, the total amount of funding cuts each chamber had approved in floor action on the bill would be averaged.(1) The discretionary caps would be reduced by this average amount for the fiscal year in question, as well as for all succeeding years for which a cap has been established. These cap reductions would be instituted even if one house had approved a cut by a narrow margin and the other house had decisively rejected it. As a result, one house's decision to cut a bill would force the other house to lower total appropriations without any concordance from that other chamber through normal conference procedures.
In addition, the lock-box mechanism could lead the discretionary caps to be reduced by more than the amount needed to "lock away" savings created by cutting a particular project. Amendments reducing funding for an appropriations bill would result in a reduction in the discretionary caps not just for the fiscal year covered by the appropriations bill but for each fiscal year after that for which there is a statutory cap. If appropriations are cut for a one-time project -- say, providing fewer funds for a NASA space-shuttle procurement or a particular construction project -- future cuts in other programs would be required.
(Ironically, one effect of this provision might be to make it more difficult to reduce low-priority spending, an effect that is the opposite of what the bill's sponsors seek to achieve. Suppose an amendment to cut a big-ticket item, such as a NASA procurement, is offered. Those who favor the cut are likely to fall into two groups -- those who want to use the savings to shrink government and those who want to shift the funds to other areas. Under current rules, both groups will join to vote for the cut. Under the procedures H.R. 853 would establish, the two groups may divide. If the amendment making the cut places the savings in the "lock-box," many in the "reorder priorities" faction may oppose it, as it will shrink the overall resources available for discretionary programs. If the amendment does not place the savings in the lock box, the "cut government spending" faction may oppose it. The result could be that fewer amendments to cut low-priority spending are approved and more, rather than less, of the status quo is maintained. If that occurred, the provision would retard efforts to reorder budget priorities.)
Using Discretionary Cuts to Finance Permanent Tax Cuts or Entitlement Increases
H.R. 853 also would enable tax cuts and entitlement expansions to be financed by reductions in the discretionary spending caps, since it would allow non-Social Security surpluses to be used for tax cuts and entitlement increases, and reductions in the discretionary caps would enlarge these surpluses. Allowing tax cuts and entitlement expansions to be financed by reducing the discretionary caps raises several concerns.
The discretionary caps typically are set for only a few years at a time; currently, the caps are in place through 2002. Lowering the caps would thus assure savings for only several years. Tax cuts and entitlement expansions, by contrast, are usually permanent and often grow in cost over time. Allowing policymakers to pay for permanent tax cuts and entitlement expansions with reductions in the discretionary caps that provide short-term savings could lead to significantly smaller surpluses or larger deficits in years to come. In addition, because reducing the discretionary caps for future years does not itself entail cutting specific programs, it could become too easy for policymakers to pay for popular tax cuts or entitlement increases by lowering the discretionary caps.
Moreover, because the effects of reducing the discretionary caps are not felt immediately, these caps could be lowered by unrealistic amounts to pay for tax cuts, with the result that the caps subsequently are raised back up and the anticipated savings not secured.
The savings in discretionary programs assumed as part of the 1997 budget agreement may be a case in point. The 1997 budget agreement instituted caps to keep discretionary spending at a virtual freeze over five years, requiring substantial reductions in discretionary programs in inflation-adjusted terms, with these reductions concentrated in the last several years of the five-year budget period. It now appears the caps are unrealistically tight and will probably be raised.
Exacerbating this problem, the bill specifies that in determining the size of the budget surplus -- and hence the amount available for tax cuts and entitlement increases -- a "baseline" must be used that assumes discretionary spending is frozen in all future years for which a statutory cap is yet to be established, with no adjustment for inflation. This represents a departure from current practice under which baseline levels for discretionary spending for future years without a cap are set equal to the prior year's levels, with an adjustment for inflation. Under H.R. 853, the baseline would assume large reductions over time in the levels of services that discretionary programs would provide. For example, using CBO's inflation assumptions, a baseline that contains no adjustment for inflation would assume a 12 percent cut in service levels after five years and a 23 percent cut by the tenth year.(2)
Altering the baseline procedures that have been in place for nearly a quarter of a century and eliminating inflation adjustments in projecting discretionary spending levels, as H.R. 853 does, will make budget surpluses look much larger and hence enable tax cuts to be substantially bigger. This one provision of the bill would artificially swell the non-Social Security surplus by more than $436 billion over the next 10 years. (See Table 1.) This could lead to much larger tax cuts and entitlement increases that, in turn, could lock in frozen or otherwise low levels of discretionary spending, necessitating substantial reductions in the levels of service that discretionary programs provide, since the tax cuts or entitlement expenses would have consumed the resources needed to support discretionary appropriations at a more adequate level.
In fact, under H.R. 853, if tax cuts had been approved that consumed the on-budget surplus, subsequent action to raise the discretionary caps to facilitate the passage of appropriations bills could trigger a sequester of Medicare and other entitlement programs. This would make it very difficult to raise the discretionary caps if the caps proved excruciatingly tight. In short, H.R. 853 could lead to deep cuts in discretionary programs.

H.R. 853 would be likely to lead to lengthy delays in action on appropriations bills. The appropriations committees would be prevented from sending appropriations bills to the House floor until work on the budget resolution had been completed. By contrast, current budget rules allow the House Appropriations Committee to send appropriations bills to the floor if action on the budget resolution has not been concluded by May 15.
In addition, the bill would change the budget resolution from a concurrent resolution to a joint resolution. Getting the two houses of Congress to agree on a budget resolution has often proved to be a lengthy process even when the same party controls both houses. Developing a budget resolution that also must win the President's approval and signature, and would have the force of law, almost certainly would be a lengthier undertaking. (H.R. 853 would allow a concurrent resolution to be used instead of a joint resolution if a joint resolution had passed Congress and been vetoed by the President.)
H.R. 853 consequently would make the process of passing a budget resolution more difficult and time consuming, while barring appropriations bills from coming to the House floor until the budget resolution had been approved, regardless of how long that might take. In many years, floor action on appropriations bills probably would not be able to commence until late in the year. In years in which budget agreements are delayed, the appropriations committees could lose months of valuable time and find themselves under great strain to put together and pass bills in compressed timeframes late in the year.
H.R. 853 would bar action on appropriations bills prior to approval of a budget resolution even if statutory caps are in place on discretionary spending. Yet such caps make the budget resolution largely superfluous insofar as discretionary spending levels are concerned. When caps are in place, there is little reason to delay appropriations actions for long periods until a budget resolution is adopted; most budget resolutions simply adopt the discretionary caps already in law. This aspect of the legislation seems particularly ill-advised.
Automatic CR
H.R. 853 also could make it more difficult in another respect to pass appropriations bills. It would establish an "automatic continuing resolution" that would maintain funding at the prior year's level for programs in appropriations bills not enacted when a fiscal year commenced. The automatic CR would not expire after a few weeks or months, but would last for the full fiscal year unless superseded by passage of the appropriations bill in question.
Although the automatic CR provision is intended to avert government shutdowns, its principal effect could be to make it more likely that Congress would fail to work out agreements on controversial appropriations bills because a year-long CR would kick in automatically. The fact that the automatic CR could remain in effect for a full year, rather than expiring after a few weeks as most current CRs do, would ease pressure to work out agreements on regular appropriations bills. Moreover, the automatic CR provision could encourage minority Senate factions of 41 or more Senators to use filibusters to block appropriations bills to which they objected, since doing so would not threaten to disrupt government operations.
The result could be that automatic CRs would begin to supplant some appropriations bills. If so, the effect would be unfortunate. Relying upon automatic CRs rather than passing regular appropriations bills would reduce government efficiency and effectiveness since it would keep Congress from addressing changing priorities. Funding levels for programs covered by automatic CRs would be stuck at the prior year's level rather than increased for some programs and decreased for others to reflect changes in need. Permanent CRs frustrate efforts both to fund promising new initiatives and to pare back less-effective, outdated, and less-important programs. If this provision of the H.R. 853 led to more reliance on CRs and fewer enacted appropriations bills, the status quo would be reinforced at the expense of more responsive and effective government.
Effects on Social Security Reform
The "pay-as-you-go" budget rules currently in place require that entitlement increases and tax cuts be paid for with reductions in other entitlement programs or revenue-raising measures. These rules apply whether the budget is in deficit or surplus. Enacted in 1990, the pay-as-you-go rules have played a large role in eliminating deficits and, over the past year, in preserving projected surpluses.
H.R. 853 would alter these rules to allow policymakers to use non-Social Security surpluses to finance tax cuts and entitlement increases. Offsetting tax increases or entitlement reductions would not be needed. Although a provision of this nature may ultimately make sense, enacting it now could make it more difficult to reform Social Security and Medicare.
Plans to restore long-term Social Security and Medicare solvency may require more resources than the Social Security surplus itself provides; some temporary general revenue transfers from the non-Social Security surplus to the Social Security and/or Medicare trust funds may be necessary to fashion solvency legislation that can pass. If action is taken to alter budget rules so the non-Social Security surplus can be consumed by tax cuts and entitlement increases before legislation restoring Social Security and Medicare solvency is approved, resources that may prove necessary for solvency legislation may disappear. That could make it more difficult to secure agreement on Social Security and Medicare legislation. (It also could mean that whatever Social Security and Medicare solvency legislation ultimately is enacted would have to contain larger benefit reductions than might otherwise be the case, because resources that could have been used to bolster the trust funds would be gone.)
Sequesters If Surpluses Do Not Materialize
This provision of the bill also poses another problem. Projected surpluses in the non-Social Security budget would essentially be used as contingent offsets for tax cuts or entitlement increases. If the surplus for a fiscal year subsequently turned out smaller than had been projected, the tax cut or entitlement expansion financed from the projected surpluses would no longer be considered to have been fully financed. To secure the needed financing, a sequester that cut Medicare and other various entitlement programs (including guaranteed student loans, the child support enforcement program, the social services block grant, farm price supports, and crop insurance, among others), would be triggered unless Congress and the President acted swiftly to fill the financing hole by cutting entitlement programs, raising taxes, or lowering the discretionary caps.
This provision of the H.R. 853 poses dangers to Medicare and various other entitlements. Policymakers would pass permanent tax cuts and/or entitlement increases based on projections of surpluses that could prove too optimistic. CBO and OMB deficit and surplus projections have been off by large margins in recent years, underestimating deficits substantially in some years and overestimating deficits or underestimating surpluses in others. If this provision of the bill becomes law and paves the way for large tax cuts this year, but surpluses subsequently turn out much smaller than current projections assume, Medicare and other entitlement programs could face large across-the-board cuts unless Congress acted swiftly to pass deep program reductions or sizeable tax increases.
For example, CBO's January 1999 forecast shows a $63 billion non-Social Security surplus in 2004. Congress might pass a tax cut that costs $63 billion in 2004 without any offsets and assume the surplus would cover it. Suppose that when 2004 arrives, however, the non-Social Security surplus for that year is only $5 billion (not counting the effects of the tax cut). CBO deficit and surplus estimates made five years in advance have, on average, been off by more than that amount.(3) If this occurred, the President would have to order an across-the-board cut of $58 billion unless Congress passed legislation cutting programs or raising taxes by that amount.
A $58 billion sequester would be larger than the biggest first-year savings ever considered in any congressional budget plan of the last two decades. It would cut Medicare provider payments by four percent and entirely eliminate a number of programs, including farm price supports, crop insurance, the Social Services block grant, and payments to states for the child support enforcement program.
This would be particularly problematic for another reason as well. One of the most common reasons a surplus projection can turn out to be too high is that the economy has slowed since the projection was made. During economic slowdowns, revenues are lower than forecast, while expenditures for unemployment insurance, food stamps, and other programs are higher.
Most economists agree that cutting spending or increasing taxes when the economy is weak can push a faltering economy into recession. This, however, is precisely what would be required under H.R. 853 if a large tax cut or entitlement increase were enacted on the basis of projected surpluses but the surpluses failed to materialize because the economy weakened. Congress would have to raise taxes or cut spending -- or an automatic across-the-board spending cut would occur -- while the economy already was heading south.
This feature of H.R. 853 would essentially resurrect one of the components of the 1985 Gramm-Rudman-Hollings law most responsible for that law's failure. The Gramm-Rudman-Hollings legislation established fixed deficit targets, enforced by across-the-board cuts if the targets were missed. It ignored the fact that because deficits swell when the economy slows and for other reasons beyond policymakers' control, the law required deepening cuts as the economy weakened. As a result, large sequesters would threaten, especially when the economy could least absorb them. Since Congress and the President could not tolerate large cuts when the economy weakened or when deficit targets were missed by large margins for other reasons beyond policymakers' control, they would engage in large-scale budget deception to make it appear as though deficit targets would be met when everyone knew otherwise, and ultimately, when all else failed and crisis loomed, they would change the targets. Eventually, the unsuccessful Gramm-Rudman-Hollings process was replaced with the much more realistic and successful procedures the Budget Enforcement Act of 1990 established. The BEA has maintained and enforced fiscal discipline without requiring fiscal retrenchment when the economy weakens or deficit forecasts become more adverse due to factors that policymakers cannot control.
The problems that this feature of H.R. 853 could cause would not be limited to periods when growth was slowing. For example, tax-cut legislation could turn out to cost more than projected because inventive tax lawyers and corporate finance departments found ways to create tax shelters Congress had not intended. If tax-cut legislation turned out to cost more than forecast and hence was not fully offset, H.R. 853 could trigger a sequester of Medicare and other entitlements. The sequestration would not touch the tax provisions that had caused the problem.
CBO and OMB forecasts of future surpluses also could prove too optimistic for a number of other reasons. CBO has cautioned that its surplus forecasts may be off by large amounts if revenues grow more slowly than it has forecast. Analysts do not fully understand why revenues have grown more rapidly than projected in recent years, and they consequently do not know the extent to which the factors that have caused this unexpected revenue growth are temporary or permanent. Revenue growth in future years could be either lower or higher than CBO currently projects and by substantial amounts. If revenue growth turns out to be significantly lower but the projected surpluses have been used to finance large tax cuts and other expenditures, as H.R. 853 would allow, deficits in the non-Social Security part of the budget would threaten, and large sequesters would loom.
Similarly, a drop in the stock market would result in lower-than-expected revenue collections, since less would be collected in capital gains taxes. That, too, could trigger a large sequester of Medicare and other programs.
CBO this year devoted a full chapter of its annual report on the budget and the economy to the uncertainty of its projections. It warned that "considerable uncertainty" surrounds its budget estimates "because the U.S. economy and the federal budget are highly complex and are affected by many economic and technical factors that are difficult to predict. Consequently, actual budget outcomes almost certainly will differ from the baseline projections..." (4) CBO reported that if its estimate of the surplus for 2004 proves to be off by the average amount that CBO projections made five years in advance have proven wrong during the past decade, the forecast for 2004 could be too high or too low by $300 billion.
This is a reason for exercising caution in the use of projected on-budget surpluses and not enacting changes in budget rules that allow the projected surpluses to be used in full for tax cuts and entitlement expansions. If large tax cuts or entitlement expansions are passed but surpluses of the magnitude projected do not materialize, H.R. 853 could lead to large sequesters of Medicare and certain other mandatory programs, large cuts in other parts of the budget, or -- perhaps most likely -- changes in law to evade these requirements, with the result that deficits would return.
Emergency Spending Procedures
H.R. 853's provisions to change procedures relating to emergency spending also warrant mention. There is broad agreement that reforms are needed in this area, and many of H.R. 853's emergency spending provisions seem useful. But these provisions also include several questionable changes. Of greatest concern, the emergency and the PAYGO provisions of H.R. 853 seem inconsistent; under the bill, emergency spending could trigger a sequester of Medicare and other entitlements.
H.R. 853 would establish an emergency reserve fund, funded within the discretionary caps. When a discretionary spending item is designated an emergency, funding would come from the reserve fund. (The amount placed in the reserve fund would be based on a historical average of the annual levels of emergency spending in recent years, which would be about $9 billion a year. If, in a given year, emergencies required more money than was available in the reserve, the budget committees could agree to exempt the additional funding from the caps.) This provision would not take effect until the discretionary caps are raised so that the caps could be set at a level that takes the reserve fund into account.
While these provisions seem reasonable, they do not mesh with the provisions of the bill that can trigger sequesters if on-budget deficits threaten to return. Suppose the projected surpluses have been used for tax cuts and some spending increases, and a major disaster or foreign military involvement occurs that requires emergency spending beyond the amount in the reserve. Congress could agree to designate the additional disaster or defense spending as emergency spending, but because this spending would result in a deficit, a sequester of Medicare and other entitlements would be triggered.
The bill also appears somewhat too restrictive. To be considered an emergency, five criteria would have to be met. Two of these criteria are that the emergency be both "sudden, which means quickly coming into being or not building up over time" and "unforeseen, which means not predicted or anticipated as an emerging need." The requirement that an emergency be both sudden and unforeseen would appear to bar emergency appropriations that are intended to address problems that were foreseen or developed gradually but turn out to be considerably more severe or long-lasting than had been anticipated, such as, potentially, needs for additional funding for peace-keeping in Bosnia. If problems such as these did not meet the new definition of emergency, other discretionary programs would have to be cut when urgent funding needs for matters such as these arose.
The bill also would accord unusual power in implementing its emergency provisions to the Budget Committees. Whenever any committee approved legislation that sought to designate an item as an emergency, the Budget Committees would determine whether the item met the definition of emergency and consequently could receive funding from the emergency spending reserve. Moreover, it would be very difficult for a floor amendment to be offered to fund an emergency need; H.R. 853 includes no mechanism to handle emergencies through a floor amendment. A proposal to offer such an amendment on the House floor would generally trigger a point of order.
Federal Insurance Programs
Having been critical of a number of the provisions of H.R. 853, let me indicate support for the bill's proposals to change the accounting of federal insurance programs. The federal budget currently treats federal insurance programs (such as flood, pension, crop, and deposit insurance) under cash-based accounting methods. Under these methods, the government is credited with revenue at the time the government collects insurance premiums and is charged with expenditures at the time the government makes claim payments. Under the accrual-based accounting methods that H.R. 853 would establish, instead of recording the flow of cash each year, the budget would record the risk that the government ultimately would have to make payments not offset by the premiums it collects.
The procedures H.R. 853 would establish would reflect the government's liabilities at the time the government assumes them. The expected net losses the government would incur over the life of an insurance contract would be recorded as a cost at the time the contractual arrangement was made. This would help policymakers understand the true costs of policies affecting government insurance programs.
There are some important concerns about how OMB and CBO would estimate the expected net losses that result from insurance contracts. For this reason, this change in accounting methods would be phased in over five years, and studies would be conducted by OMB, CBO and GAO during the phase-in period. Two years after the accounting methods were fully implemented, they would expire, and Congress could decide whether they had been sufficiently successful to continue their use. This seems a prudent course to follow.
Conclusion
H. R. 853 contains some improvements in the budget process, such as its provisions reforming the accounting of federal insurance programs. Unfortunately, its deleterious aspects are serious and substantially outweigh its beneficial aspects. For these reasons, I would strongly recommend against enacting this legislation.
3. See Congressional Budget Office, The Economic and Budget Outlook: Fiscal Years 2000-2009, January 1999, p. 81.
4. Congressional Budget Office, The Economic and Budget Outlook: Fiscal years 2000-2009, January 1999, p. 81.