May 23, 2002

THE GREGG-FEINGOLD BUDGET PROCESS PROPOSAL
Problems with the Proposal Outweigh Its Good Points

by Richard Kogan and Robert Greenstein

PDF of this report

Related reports

If you cannot access the files through the links, right-click on the underlined text, click "Save Link As," download to your directory, and open the document in Adobe Acrobat Reader.

On May 7, Senators Judd Gregg and Russ Feingold introduced the Budget Enforcement Act of 2002, S. 2465. This analysis describes the main provisions of the bill. It finds that two of its aspects are particularly problematic: its dollar limits on annual appropriations are likely to be unrealistically low and would result in significant program cuts in coming years, and its rule against causing or increasing a non-Social Security deficit through the enactment of tax or entitlement legislation is not well designed and could have a number of adverse side-effects.

Appropriations limits or "caps"

Assuming the President's budget proposals for defense, homeland security, and international affairs are adhered to, the Gregg-Feingold appropriations caps would require significant cuts in domestic appropriations.

Modified "Pay-As-You-Go" Rule

Contrary to what one might hear, the Gregg-Feingold proposal would not extend the current "pay-as-you-go" rule, which has been in statute since 1990 and was effective until three years ago in controlling the cost of tax and entitlement legislation. Instead, the Gregg-Feingold proposal would significantly alter the current pay-as-you-go rule by allowing new tax cuts and new entitlement increases to be enacted without offsets, as long as budget projections in future years do not show deficits outside of Social Security.

The Mechanics of Pay-As-You-Go, as Modified by Gregg-Feingold

Suppose Congress wished to enact a tax cut or entitlement increase. If Congress "paid for" these costs with offsetting entitlement cuts or tax increases, no problems would arise. Under Gregg-Feingold, however, Congress would not be required to pay for these costs; rather, projected surpluses outside of Social Security could be used to cover the costs.

Suppose Congress enacted a tax cut and did not pay for it, based perhaps on an OMB projection that future surpluses outside of Social Security would occur, or based on wishful thinking. In such a case, each autumn, as Congress adjourned, OMB would estimate whether the cost (in the coming fiscal year) of this previously enacted tax cut had in fact been covered by a surplus. That is, OMB would estimate whether the coming fiscal year would still be in surplus (not counting Social Security) even accounting for the tax cut.

If OMB estimated a surplus for the coming fiscal year, all would be well. But if it estimated a deficit, various entitlement programs would be subject to automatic, across-the-board cuts to make up for the costs of the tax cut that had never been paid for.

Each year, through 2007, OMB would again check to see if the costs of the tax cut in the coming year had been covered by a surplus. If not, various entitlements would be subject to another year of automatic cuts.

This is essentially what happened with the large tax cut enacted last year. It seems unwise to create budget rules under which such a fiscal mistake can be repeated.

Furthermore, the Gregg-Feingold proposal could lead to budget cuts or tax increases during economic slowdowns in order to avoid sequestration, which would deepen downturns and increase hardship.

The Gregg-Feingold proposal also makes other, more minor changes in the budget process. One such change — the extension of Senate rules requiring 60 votes for measures that breach a congressional budget resolution — is beneficial from the standpoint of fiscal discipline. Another change equalizing the treatment of defense and non-defense emergencies appears to be largely cosmetic.

 

Appropriations Limits, or "Caps"

Selected Provisions of the
Gregg-Feingold Bill

Appropriations limits, or "caps"
Non-Social Security deficits and the modified "pay-as-you-go" rule
Senate enforcement of congressional budget plans
The "emergency" exception

Appropriations limits, usually called "discretionary caps," apply to non-entitlement programs funded annually by the Appropriations Committees. These programs include the Defense Department, most education programs, highways and other transportation programs, scientific research, law enforcement, veterans hospitals and some public health programs, NASA, low-income housing assistance, environmental protection, national parks and other natural resources, general government costs, and some programs providing job training, social services, and nutrition assistance. From 1985 through 2000, these programs as a group declined substantially both as a share of the economy and as a share of the federal budget.

Appropriations caps set maximum levels for these annually funded programs. Such caps were first enacted for fiscal years 1988 and 1989. Caps were re-established for fiscal years 1991-1995, this time enforced by "sequestration," which requires that the President cut appropriations across the board if Congress has enacted appropriations that exceed the caps. Caps and the sequestration process were extended twice more, although the caps were raised slightly in 1998, evaded through gimmicks in 1999 and 2000, and raised substantially in 2001 and 2002. Both the caps and the sequestration mechanism to enforce them are due to expire at the end of fiscal year 2002.

The Gregg-Feingold proposal would establish appropriations caps for fiscal years 2003 through 2007. It also would continue presidential sequestration as the tool of enforcement.

During some past years, separate appropriation caps have been established for various "categories" of appropriated programs, e.g., for defense, international, and domestic programs. The Gregg-Feingold proposal does not re-establish separate caps for defense and domestic programs but covers them both under a single cap.(1)

The overall dollar levels of the Gregg-Feingold caps equal the dollar levels in the congressional budget plan reported April 11 by the Senate Budget Committee. These levels may be unrealistically tight in 2003. They grow substantially tighter in subsequent years and almost certainly are unrealistic for those years.

In the abstract, these aggregate caps may appear generous; total funding would be allowed to grow faster than inflation in 2003, by almost $36 billion or 4.9 percent. But the reality is very different. The President has proposed large increases in defense and homeland security and some increases in the international affairs budget. These increases are likely to be fully funded this year. If they are, the amount of funding that would remain available for domestic programs would be less than needed to cover inflation in 2003. In other words, domestic appropriations (outside homeland security) would have to be cut.

The largest problems would occur in years after 2003. If the President's current budget plan for defense, homeland security, and international affairs is adhered to in future years, the funding cuts for domestic appropriations would reach 10 percent by 2007, relative to the 2002 level adjusted only for inflation (i.e, relative to CBO's baseline). The cuts would total $82 billion over the five-year period from 2003 through 2007.

How Restrictive are the Proposed Gregg-Feingold Funding Limits?
Budget authority in billions of dollars
2003 2007 5-year total
Proposed funding limits, total 768.1 841.0 4,021.9
Bush proposed budget for defense, homeland security, and international affairs 441.6 511.5 2,353.6
Amount remaining for domestic programs 326.5 329.5 1,668.4
Amounts remaining for domestic programs $ % $ % $ %
2002 levels adjusted for inflation (CBO -6 -2% -38 - -82 -5%
2002 levels adjusted for inflation and -9 -3% -54 - -128 -7%
2002 levels adjusted for GDP growth -15 -4% -90 - -228 -
NOTE: columns may not add due to rounding.

Moreover, the size of the potential domestic cuts will be larger than just mentioned if funding for defense, homeland security, or international affairs in years after 2003 exceeds the amounts shown in the President's budget, since that would leave even less room for domestic programs within the caps. This could well happen. The President's defense figures do not include funding for the war on terrorism after 2003; those costs are likely to be large. Additionally, the President's budget does not reflect any of the $5 billion increase in foreign assistance for 2004 that the President announced in Monterey, Mexico, on March 14.(2)

The table above also shows that the amount remaining for domestic programs is even further below a baseline that adjusts for both inflation and population, i.e. that keeps the inflation-adjusted, per-person level of goods and services provided by annual appropriations constant over time. Many or most domestic programs can be expected to grow with the growth in the U.S. population, from highways to law enforcement to education to public health. This standard — covering inflation and population growth — is salient for two reasons. First, it was the standard that Governor Bush used when running for President in 2000; he said it was the "honest" way to evaluate budget growth.(3) Second, from 1986 on, the average growth rate of funding for domestic appropriations has exceeded the combination of inflation and population growth. (It has done so in 10 of the last 14 years.) It is not realistic to assume that, starting this coming year, it will fall well below this standard for five straight years. The table shows that, assuming the levels in the President's budget for defense, homeland security, and international affairs, the Gregg-Feingold caps would result in domestic funding levels that fall $9 billion below this standard in 2003 and $128 billion below this standard over five years, with the gap reaching 14 percent by the fifth year.

Finally, the table shows that, under the Gregg-Feingold caps, the amount remaining for domestic appropriations would fall significantly as a share of the economy. By 2007, this amount would reach an historically low level, probably the lowest level since 1963.(4) It seems highly unlikely that policymakers would allow domestic appropriations to contract this much as a share of the economy. Funding for domestic appropriations has been roughly constant as a share of the economy since 1985.

 

Non-Social Security Deficits and Legislation Affecting Taxes and Entitlements

Material distributed by Senators Gregg and Feingold states that their proposal "would reinstate and extend the pay-as-you-go discipline that controls entitlement spending and tax law changes. Points of order and the threat of across-the-board cuts would continue to provide enforcement." But their proposal is actually very different from the existing pay-as-you-go system.

First, in many cases, their proposal would not require that tax cuts or entitlement increases be paid for. Their proposal would enforce this rule only when OMB projected in the fall that the budget for the coming year would be in deficit outside of Social Security. If optimistic OMB projections showed future surpluses (as, for example, they did 15 months ago) or if Congress was willing to gamble that budget forecasts would improve, Congress could feel free to enact tax cuts or entitlement increases without paying for them at all.

The Gregg-Feingold rule thus could entice Congress into enacting additional tax cuts or entitlement increases that are not paid for. If so, Congress would then find itself at the mercy of economic and budget estimates made in subsequent years. If the projections made in subsequent years were rosy enough to show a surplus, no automatic entitlement cuts would occur. But if forecasts became less optimistic and showed deficits outside of Social Security, various entitlement programs would be subject to automatic cuts. In this way, Congress would be at the mercy of estimates made by the Administration, estimates that often shift substantially or could be manipulated for political reasons. In this respect, the Gregg-Feingold proposal is closer to the discredited Gramm-Rudman-Hollings law of the late 1980s.

The existing pay-as-you-go rule

To understand the critical difference between the existing but expiring pay-as-you-go rule and the Gregg-Feingold proposal, we start by examining the existing pay-as-you-go rule, which was initially designed to protect the bipartisan deficit reduction agreement of 1990 against backsliding. The existing pay-as-you-go rule requires that any tax cuts or entitlement increases be fully offset, or "paid for," by other tax increases or entitlement cuts. In essence, the pay-as-you-go rule is a rule of budget neutrality for tax or entitlement legislation.

A key aspect of the pay-as-you-go rule (and of the appropriations caps, for that matter) is that the size of the actual or estimated budget deficit or surplus does not matter. Adherence to the rule of budget neutrality is required no matter how dim or rosy the overall budget estimates seem. This requirement means that Congress does not have to raise taxes or cut entitlements during an economic slowdown, and in return does not get to "spend" surpluses during boom times.

The Gregg-Feingold pay-as-you-go rule

The Gregg-Feingold rule extends the existing pay-as-you-go rule for another five years but substantially changes its effect by adding an exception: "Notwithstanding any other provision of law, there shall be no sequestration under this section for any fiscal year in which a surplus exists." The Gregg-Feingold proposal defines the term "surplus" to exclude Social Security for this purpose — that is, it refers only to the non-Social Security budget. For ease of discussion, the remainder of this analysis will use the term "surplus" to mean a surplus outside of Social Security.

Under the Gregg-Feingold version of the pay-as-you-go rule, Congress would not have to wait until CBO or OMB again projected surpluses. Rather, Congress could enact new tax cuts or entitlement increases without paying for them whenever it wanted; Congress could hope or gamble that when the time came for OMB to see whether sequestration was needed — a determination OMB would make each fall — OMB would honestly or miraculously project that the budget would be in surplus for the coming year, precluding the need for sequestration.

To see how this new rule might operate, let's work through a hypothetical example, in three steps.

First, suppose we were starting over from scratch, with the Gregg-Feingold proposal in place. Now imagine that Congress enacted a tax cut costing $0 billion in fiscal year 2003, $21 billion in 2004, $24 billion in 2005, $27 billion in 2006, and $30 billion in 2007. (These costs are similar to those requested by President Bush in February for additional tax cuts, beyond those intended for "economic stimulus.") Because these new tax cuts would not have been paid for, OMB would be required to record these costs on its "pay-as-you-go" ledger.

Second, when Congress adjourned in the autumn of 2003, OMB would have to decide whether a sequestration of selected entitlements was required. OMB would check its "pay-as-you-go" ledger for the fiscal year that has just started — fiscal year 2004 — and find $21 billion in 2004 costs on the ledger, which had been caused by the above tax cuts. Under the old pay-as-you-go rules, those costs would trigger a sequestration because they had not been offset. But under the proposed Gregg-Feingold exception, whether a sequestration is required would depend on the OMB budget projections.

Third, OMB would project whether the fiscal year 2004 budget would be in surplus or deficit. If OMB projected a 2004 deficit of, say, $100 billion, there would be a $21 billion sequestration of various entitlement programs to offset the $21 billion in tax cuts that had not been paid for. If OMB projected a 2004 surplus, there would be no sequestration.

In practice, OMB probably would not wait until Congress adjourned to inform Congress of its projection of the surplus or deficit for the coming fiscal year.(5) This means that OMB or Congress would be able to use the new rules to achieve political purposes. For example —

Differences between the existing pay-as-you-go rule and the Gregg-Feingold rule

In summary, the Gregg-Feingold proposal differs from the existing pay-as-you-go procedures in four significant ways.

This is a risky approach. Last year's tax cut was justified in part on the basis of projected surpluses that, only one year later, seem naively optimistic. The obvious risk is that the economic forecast on which a tax cut or entitlement increase is justified will prove to have been too optimistic. Another risk is that sudden, unexpected funding needs will arise, such as those after September 11 or those caused by major natural disasters. Congress can exempt such costs from the caps by designating them an emergency, but these costs will still be real and could still turn a surplus into a deficit.

As this discussion indicates, the Gregg-Feingold proposal resembles the Gramm-Rudman-Hollings budget process regime of the second half of the 1980s, when a fixed target for was established for each year's budget deficit and sequestration was supposed to be triggered if OMB projected that the target for the coming year would be missed. Under the Gregg-Feingold version of the pay-as-you-go rule, the fixed target would be a balanced budget outside of Social Security.

The late 1980s was marked by official over-optimism in budget estimates (to the point where OMB's figures and figures in congressional budget plans were routinely ignored as biased), increased use of timing shifts to solve one year's budget problems at the expense of another year's budget, and a reliance on across-the-board appropriations cuts as a last resort even when appropriations bills met their other targets. Gregg-Feingold could result in the return of these same unfortunate budget practices.

 

Senate Enforcement of Congressional Budget Plans

Under the Congressional Budget Act, it is against the rules of the House and Senate to consider legislation that would spend more than the congressional budget plan (or "budget resolution") allocates to each committee of Congress. Likewise, it is against the rules to consider tax cuts that would exceed the amount of tax cuts (if any) called for in the congressional budget plan.

In the Senate, these prohibitions can only be waived by the vote of 60 senators, but that requirement expires on September 30, after which these rules can be waived by majority vote. The Gregg-Feingold proposal would extend the need for 60 Senate votes to breach the limits in a congressional budget plan for another five years. Of all the proposals in the Gregg-Feingold budget proposal, this may do the most to promote fiscal discipline.

 

The "Emergency" Exception

Under current law, amounts that are designated by the Congress and independently by the President as emergencies are exempt from all budget limits, whether limits on tax cuts or program increases and whether enforced by congressional points of order or OMB sequestration. However, it is against the rules of the Senate to designate a cost as an emergency. This Senate prohibition against an emergency designation can be waived by majority vote in the case of defense emergencies but requires the vote of 60 Senators in the case of nondefense emergencies.

The Gregg-Feingold proposal would eliminate the apparent special advantage given to defense emergencies — any emergency designation would require 60 votes. In practice, the distinction has turned out to be artificial. In fiscal 2000, the House and Senate Appropriations Committees made room for increased domestic spending by designating more than $7 billion in routine defense funding for operations and maintenance as emergency spending. The Committees did not increase defense funding by that amount; they merely designated as an emergency that amount of funding, which had already been agreed to by House and Senate conferees on the defense appropriations bill. This designation allowed the House and Senate Appropriations Committees to re-allocate $7 billion from the defense subcommittees to other subcommittees, which could then use the additional room within their allocations without having to rely on an emergency designation. Given this history, this Gregg-Feingold proposal is more accurately viewed as tightening emergency designations generally than as eliminating favored treatment for defense spending.

 

Conclusion

The two main features of the Gregg-Feingold proposal both are problematic. The appropriations caps are very likely to be too tight, especially in the outyears. If Congress adheres to the President's budget plan for defense, homeland security, and international affairs, the amount remaining under the caps for domestic programs would constitute a cut in every year, with the cuts becoming deeper with each passing year. Moreover, it would be unfortunate for presidential or congressional budgets to assume an unrealistically low level of domestic funding in the outyears because that would make tax cuts or entitlement increases appear more affordable than they really are.

In addition, the Gregg-Feingold version of the pay-as-you-go rule is very different from the existing rule. The new version would provide an exception — the rule would not have to be enforced in any year in which OMB projected a surplus outside of Social Security. This exception could tempt Congress and the President into gambling on the appearance of future surpluses and being cruelly disappointed — and either suffering major, automatic cuts in selected entitlements such as Medicare and farm programs or resorting to extensive gimmicks or rosy estimating scenarios to pretend the requirements had been met (or engaging in last-minute cuts in appropriated programs to make a projected deficit disappear). The history of the Gramm-Rudman-Hollings rule of the late 1980s, a somewhat similar rule, suggests that gimmicks, rosy scenarios, and appropriations cuts are indeed the likely outcomes. History also suggests that under such a rule, OMB can use its authority to issue deficit or surplus projections as a club against Congress if it wants to force one of these outcomes, such as additional, last-minute appropriations cuts.


End Notes:

1. Currently, there exist some minor, separate categories: highways, mass transit, and some conservation programs. The Gregg-Feingold proposal leaves these minor categories in place (and the special treatments associated with some of them) until their scheduled expirations in the next few years.

2. The figures for defense, homeland security, and international affairs may be understated in two other ways as well. First, they do not include the amounts in the President's budget to increase agency "accrual" payments to the civil service retirement trust fund. Second, they are calculated on the assumption that funding for homeland security grows only with inflation after 2003.

3. For example, the George W. Bush for President official web site stated: "When adjusted for inflation and population, [Texas] state spending will increase by only 3.6 percent between1994-1995 and the end of the 2000-2001 biennium." Similarly, the Dallas Morning News reported: "Wednesday, [Governor Bush] said an 'honest comparison' of spending growth should take inflation and the state's increasing population into account" (October 28, 1999).

4. Data for domestic funding are not available before 1976. However, the path of expenditures (rather than funding) for domestic appropriations suggests that the levels in the table would be the lowest since 1963.

5. The Gregg-Feingold proposal does not say what economic, technical, or policy assumptions OMB must use in projecting whether the budget will be in surplus or deficit. This will give OMB considerable leeway to adopt a rosy view and determine, on that basis, that no sequestration is needed.