Chairman Goss, Chairman Horn, and Members of the Subcommittees, I am pleased to be here today to discuss proposals to provide for a deficit reduction "lockbox." My testimony will discuss how the proposals for a lockbox would change the way in which discretionary spending is controlled and will also explore some issues that the Congress may want to consider before making such a legislative change.
 

CONTROLS ON DISCRETIONARY SPENDING

Since 1990, discretionary spending has been controlled by statutory limits, or caps, on total budget authority and outlays appropriated for discretionary programs. Proposals for a lockbox would create an additional means for the Congress to reduce discretionary spending by allowing Members of Congress to offer amendments to appropriation bills that would reduce spending for specific programs and apply the resulting savings to deficit reduction.

The Budget Enforcement Act (title XIII of the Omnibus Budget Reconciliation Act of 1990) currently controls discretionary spending through statutory limits on budget authority and outlays. Those limits governed three categories of discretionary spending (domestic, defense, and international) through fiscal year 1993. Since then, one cap has existed for all discretionary spending. The current caps expire in 1998, but the budget resolution recently approved by the Congress anticipates continuing the caps through 2002. Under that control system, if the cumulative effect of all appropriations for a single year would exceed the caps, the Office of Management and Budget (OMB) can order a sequestration, or an across-the-board cut, of all discretionary programs.

Actually, the caps created a statutory control on top of the controls that already existed in the budget process since the Congressional Budget and Impoundment Control Act of 1974 was passed. Under that act, the budget resolution specifies a total level of budget authority and outlays, which are then allocated to committees with spending jurisdiction. In the case of discretionary spending, the allocation is made to the Committees on Appropriations. That committee allocation, referred to as the section 602(a) allocation (after the section of the Budget Act that established it), provides a limit on the amount of budget authority and outlays that can result from action by the Appropriations Committees. The existence of the caps effectively establishes a ceiling on the level of the 602(a) allocation for discretionary spending. The budget resolution may provide less than the level of the cap for discretionary budget authority and outlays, but it may not provide more.

After the 602(a) allocations are provided to the Committees on Appropriations, those committees subdivide their allocations by subcommittee (so-called 602(b) allocations). Each subcommittee is limited by the amount of its 602(b) allocations. Bills reported out of the committee may not exceed the amounts permitted under the 602(b) allocations. Similarly, the total of all 602(b) allocations cannot exceed the 602(a) allocation. The Committees on Appropriations are, however, permitted to revise the subcommittee allocations as they choose; indeed, they have routinely done so several times each year as they have considered appropriation bills.

Currently, if a successful amendment is offered to reduce budget authority in an appropriation bill (either in committee or on the floor) and the reduction would cause that bill to come in below its 602(b) allocation, one of three things can happen. First, that money could be made available within a subcommittee's own jurisdiction to spend on other programs. Second, if the subcommittee allocations are revised, the money could be made available for other subcommittees to spend on programs within their jurisdictions. Third, if no additional funding is made available for spending on other programs, total discretionary spending would be cut by the amount of the reduction.

Consider, as an example, a successful amendment to the Veterans Affairs (VA), Housing and Urban Development (HUD), and Independent Agencies appropriation bill to reduce funding for the space station by $1 billion. Under the current process, that money may be spent on other programs within the same appropriation bill. Alternatively, the 602(b) allocations may be revised to permit the $1 billion to be spent on programs in one or more other appropriation bills. For instance, the allocation for the Department of Defense appropriation bill might be increased by $1 billion as a result of the reduction in space station funding. Finally, the Congress might choose to make the reduction without funding a corresponding increase anywhere else in the budget. Presumably, that would mean that the VA-HUD bill would come in under its 602(b) allocation, and the 602(b) allocations for other subcommittees would remain unaltered. The result would be a $1 billion reduction in total discretionary spending.

That control process for discretionary spending is one in which the decisions on the size of the total discretionary pot are made first; then decisions are made on how to allocate those funds. Usually, as a result, total appropriated spending for a given fiscal year is close to the amounts permitted under the discretionary spending limits and to the total for discretionary spending specified in the budget resolution. Thus, the appropriation process sets priorities within an agreed-on level of discretionary spending and then decides how much money goes for defense programs, health programs, justice programs, and the like.

Some Members of Congress, however, are dissatisfied with the current process. They maintain that it does not permit a majority of Members of a given body to require that a cut in an appropriation bill be used to reduce the deficit, as opposed to permitting spending in other areas. Members have been particularly concerned that successful efforts to eliminate funds for certain programs or projects did not reduce the deficit, but resulted in additional money being added for other programs at the conference stage. For that reason, they have suggested that Members be able to offer budget-reducing amendments and earmark the savings for deficit reduction by placing the amounts from the cuts into a deficit reduction lockbox.

Proposals for the lockbox vary and have surfaced recently in many contexts. The House version of the line-item veto legislation included a lockbox provision, as did title IV of the House-passed Emergency Supplemental Appropriations/ Rescissions Act for Fiscal Year 1995 (H.R. 1158). Most recently, Members have raised the issue again about the fiscal year 1996 appropriation bills. Most of those proposals would permit amendments that cut spending to be used for deficit reduction in two ways--by preventing the 602(b) allocations from being revised to reallocate funds that were reduced, and by requiring a reduction in the discretionary spending caps equal to the total amount placed in the lockbox.

Proposals vary concerning how much discretion proponents of deficit-reducing amendments would have to allocate reduced funds to other programs. Under the strictest version of the lockbox proposal (similar to that offered by Congressman Brewster to the rescission bill considered earlier in this session), funds would not be available under any circumstances for other spending. Rather, the discretionary spending limits would be reduced by the amount placed in the lockbox. Under H.R. 1162, introduced by Congressman Crapo, a proponent of a deficit-reducing amendment would have the option of transferring the amount reduced to other programs within the same bill or placing the funds in the lockbox. A third approach would give Members the flexibility to offer amendments to include not only the options of placing funds in the lockbox or using them for other programs within a subcommittee's jurisdiction, but allowing the funds to be used for other discretionary spending programs (in other subcommittees) as well.
 

ISSUES TO CONSIDER IN EVALUATING THE LOCKBOX PROPOSAL

The Congress currently has no way to guarantee that savings resulting from individual amendments to reduce discretionary spending will actually lower the deficit below the level established by the appropriation caps. For example, the incentives that now exist for the Appropriation Committees or for the Congress as a whole may discourage efforts to reduce discretionary spending below the level that is provided by the caps or the budget resolution. Subcommittees spending less than their allocation do so with the knowledge that the funds will be made available for spending by some other subcommittee. Perhaps if they could earmark funds for deficit reduction, they would have more incentive to cut programs.

The Congress, however, should consider several issues in evaluating the appropriateness of the lockbox idea. First, how necessary or desirable are additional controls on discretionary spending? Second, what is the most appropriate way to control discretionary spending? Third, what technical issues do the current proposals raise that might need to be addressed before enacting the lockbox idea into law?

Is Discretionary Spending Under Effective Control?

Most budget experts, including the Congressional Budget Office (CBO), maintain that the prime engine of growth in spending is mandatory spending programs rather than discretionary spending. That is particularly true since the caps were adopted in 1990. In fact, discretionary outlays were less than 10 percent higher in 1994 than they were in 1990 (and most of that growth occurred between 1990 and 1991). Moreover, the current caps would freeze discretionary spending in place through 1998. Indeed, as a percentage of gross domestic product (GDP), discretionary spending will have declined from 9.2 percent in 1990 to 6.8 percent in 1998, if the Congress complies with the current caps.

The budget resolution recently approved by the Congress anticipates a further ratcheting down of discretionary spending in real terms between 1998 (when the current caps expire) and 2002. The resolution would continue the freeze on discretionary outlays between 1998 and 2002. By 2002, discretionary spending would be $164 billion (approximately 25 percent) below the level in CBO's inflation-adjusted baseline (see Table 1 for a summary of the caps assumed in the budget resolution compared with those in the baseline).

Conversely, mandatory spending has been growing rapidly over the same period. Spending for entitlements and other mandatory programs, which is not controlled by the appropriation process, was almost 40 percent higher in 1994 than in 1990 and is projected to continue at only a slightly lower rate of growth through 1998. CBO projects that mandatory spending will equal 12.6 percent of GDP in 1998 compared with 10.3 percent in 1990. Medicare and Medicaid have fueled much of that rapid growth in spending. The budget resolution would slow their rate of growth but would still permit mandatory spending to increase as a percentage of GDP.

In short, the caps have proved to be an effective control over discretionary spending, but mandatory spending has continued its rapid growth. Faced with that scenario, the Congress may want to consider whether additional controls over discretionary spending are desirable or if the real focus for reform should be on the mandatory programs, which have not proved to be as tractable.

TABLE 1. TOTAL DISCRETIONARY OUTLAYS (By fiscal year, in billions of dollars)
	1995	1996	1997	1998	1999	2000	2001	2002
CBO March 1995 Uncapped Baseline*	548	556	574	592	613	635	657	679
Budget Resolution"	548	534	524	518	516	520	516	515
Budget Resolution Minus Baseline	0	-22	-50	-74	-97	-115	-140	-164
SOURCE: Congressional Budget Office.
NOTE: Numbers may not add to totals because of rounding
a. Assumes appropriations for 1996 through 2002 at the fiscal year 1995 enacted level adjusted for inflation.
b. U.S. House of Representatives, Conference Report on House Concurrent Resolution 67, the Concurrent Resolution on the Budget for Fiscal Year 1996, H. Rpt 104-159 (June 26,1995).

What Is the Appropriate Means of Controlling Discretionary Spending?

If the Congress decides that additional controls on discretionary spending are desirable, the lockbox concept and the existing method of control represent fundamentally different ways of meeting that goal. Under the existing process, the focus is on controlling spending by setting the totals at an early point in the process. Under the concept of a lockbox, decisions on the total level of discretionary spending can be revisited often on a case-by-case basis.

The existing system has relied on a set of controls by which the President and the Congress agree on multiyear enforceable targets for discretionary spending. As long as those targets are reached (and a sequestration is triggered if they are not), the deficit is reduced by the agreed amount. Additional deficit reduction could be brought about primarily either by the President and the Congress reaching a new agreement or by establishing targets in the budget resolution that would hold appropriations below the level of the cap. The appropriation process itself focuses on setting priorities within the constraints reflected by the cap, but ^normally no significant additional spending reductions result from the appropriation process. The role of the Appropriations Committees in the process has been to decide the allocation of discretionary spending among different programs, not the level of total discretionary spending.

The lockbox approach would give the Congress the flexibility to revisit decisions on the total amount of discretionary spending as each appropriation bill emerges from committee. Each time a lockbox amendment was considered, the respective House could go on record as saying that the existing caps are too generous and could advocate their reduction. The eventual reduction in discretionary spending, of course, would still depend on the outcome of the conference between the two Houses and the approval of the President. But the approach would represent a fundamentally different method of control--one in which the total amount of permissible discretionary spending would change on a week-by-week or even day-by-day basis during the period when appropriation bills were being considered.

Nothing is inherently right or wrong with those two approaches. But considering the lockbox proposal cannot be separated from deciding on how best to control the level of discretionary spending. The discretionary caps, as noted, already represent a nominal freeze in appropriated spending. Moreover, they are likely to be reduced further as a part of the effort to reach a balanced budget. If the Congress believes that further reductions in discretionary spending are desirable, it could reduce the caps even more--that is, by setting lower targets at the outset of the process. Alternatively, if the lockbox approach is selected, it would leave open the possibility--but not necessarily the certainty--that the caps may be reduced further in the future.

The Congress needs to weigh the benefits of the flexibility that would be created under the lockbox reform against the possible costs that could result because of the potential for continual political conflict over the total amount of discretionary spending. Alternatively, because the lockbox reform would leave the decision on a further reduction in discretionary spending until the future, it would provide policymakers with greater flexibility to respond to changing circumstances than the setting of the appropriation caps does. Nonetheless, changing the rules after the President and both Houses have already agreed to the appropriation limits could make it even more difficult than it currently is to reach agreement on appropriation bills. Even in the current process, the House, the Senate, and the President frequently disagree over priorities within a given total level of spending. The lockbox approach would add the possibility of frequent disagreements over the total amount of discretionary spending. That may be more conflict than the appropriation process could bear.

Another related issue concerns the ability of the Congress to set priorities for discretionary spending. As noted in the description of the lockbox concept, some approaches offer less flexibility than others in using savings from amendments. In drafting a lockbox approach, the Congress would need to decide on a number of issues. Should all spending reductions go to reduce the deficit? Should amendments be permitted to provide for offsetting increases within the same bill? Or should amendments free up funds for other discretionary spending purposes? For example, a Member who wanted to reduce funds for the space station in order to provide more funds for defense would be prohibited from doing so under H.R. 1162.

What Technical Issues Are Raised by the Proposals?

Some of the proposals for a lockbox raise several minor technical issues. Those include the necessity of creating a deficit reduction trust fund, the desirability of reducing discretionary spending limits in future years to reflect amounts placed in the lockbox, and limitations on the ability of CBO to inform the Congress unilaterally of the status of spending legislation.

A Deficit Reduction Trust Fund. Several proposals would create a special trust fund to which the amounts placed in the lockboxes from the various appropriation bills would be transferred. Those transactions would have no additional effect. The deficit reduction represented in the lockbox would result from reducing the discretionary spending limits. That adjustment would be made regardless of the status of the trust fund. The trust fund is unnecessary and could prove confusing when compared with more conventional trust funds, which exist primarily as a means of accounting for earmarked taxes that are limited in their use to specific purposes (as is the case, for example, with the transportation trust funds or the Social Security trust funds).

The Out-Year Effects of Reducing Discretionary Spending Limits. The lockbox provision attached to H.R. 1158 would require OMB to adjust future discretionary spending limits downward by the amounts placed in the lockbox, adjusted for inflation. That requirement ignores the reality that appropriation decisions made in one year do not necessarily reflect in any way the priorities and necessity for spending in succeeding years.

In some cases (such as for many agency salary and expense accounts), a cut in one year may translate into continuing savings. But in other cases, the relationship between a cut this year and savings in the future are less straighforward. Consider an effort to cut funds from a Department of Defense appropriation bill for the purchase of a submarine. Canceling funds to purchase a submarine in one year implies nothing about the need for that submarine in another year. Nor is it necessarily true that submarines would be purchased every year. Perhaps the priority in the next year would be aircraft carriers. If so, should the canceling of a submarine this year automatically imply that the Defense Department should have less money to spend on aircraft carriers next year? The appropriation process depends very much on current circumstances. Adjustments to the discretionary spending limits, then, should probably be made only one year at a time.

CBO's Ability to Inform the Congress. Finally, let me note a relatively minor point, but one that affects CBO directly. Section 4 of H.R. 1162 would require that CBO "score all general appropriation bills as passed the House of Representatives and as passed the Senate and have such scorecard published in the Congressional Record." CBO already scores appropriation bills as they pass both Houses, and we provide that information to the Budget Committees. But CBO, as a support agency of the Congress, cannot unilaterally cause anything to be published in the Record. It would seem more appropriate, therefore, for that section to be amended to specify that CBO provide the information to particular committees in the Congress (such as the Budget Committees or the Appropriations Committees), which would in turn have the information published in the Record.
 

CONCLUSION

In short, my main message is that the evaluation of the lockbox idea should hinge on two factors. First, how much should discretionary spending be constrained beyond the levels reflected in current law and anticipated by this year's budget resolution? Second, how much does such additional restraint require using the lockbox approach, as opposed to simply providing for lower discretionary caps?

Finally, perhaps resolving such issues should wait for the results of the comprehensive review of the budget process that your Committees are now undertaking. Certainly the question of what change might be made to the method of controlling discretionary spending is an appropriate focus of that review. But the lockbox idea is only one of many proposed reforms that might be considered. For example, proposals have surfaced that would require a vote of the full House and the full Senate to ratify the allocations made to the appropriations subcommittees. Other proposals would eliminate the functional categories in the budget resolution and replace them with allocations to committees and subcommittees. A consideration of the merits of the lockbox might be left until your Committees have completed their review of the process as a whole. At that time, if a change is sought in the way that appropriated spending is controlled, the lockbox approach could be considered as one option.