Before the Committee on Homeland Security and Governmental Affairs
Subcommittee on Federal Financial Management,
Government Information and International Security
Chairman Coburn, Ranking Member Carper, and members of the committee, thank you for the privilege of appearing today to discuss the federal budget process. This is an important and timely hearing; your leadership in addressing improvements to the way budgeting is conducted by the federal government is to be commended.
Goals for the Budget Process
While there has been a fairly widespread discussion of the budget process, criticism of its failings and suggestions for changes I think it is useful to begin by stepping back a bit and asking the framing question: what should be the goals of the budget process? The answer, I think, is that the budget process should:
- Identify decisions that the Congress must make, and
- Serve to implement the choices that the Congress makes.
Notably missing from this list is both “force the Congress to make decisions” and “substitute for the absence of Congressional decisions.” It is a source of understandable frustration to many that contemporary budget policy is devoid of much-needed strategies to slow the growth of mandatory spending for retiree income and health care, to take the most significant example. In these circumstances, it is tempting to impose a budget process that “solves” these problems.
But this is misguided. The budget process cannot be a substitute for policy choices by the Congress. Nor can it be expected to supplant the necessary elements of proposal, counterproposal, discussion, disagreement, and compromise that constitute the politics of policy decisions. Any such efforts to do so will be doomed to ultimately fail.
Desirable Characteristics of a Budget Process
To meet its objectives of identifying decisions and implementing choices, a budget process requires several characteristics. It should be:
- Comprehensive in scope,
- Focused on legislation and policy decisions,
- Accurate and balanced in its presentation, and
- Tied to the financial management system.
Comprehensive in scope. A first requirement of a budget process is that it covers the entire budgetary landscape. That is, there is no place for commitments that require taxpayer resources to be “off-budget” in any real sense. Equally importantly, the budget process should be comprehensive in a second sense—identifying all the decisions that need to be made.
Focused on legislation and policy decisions. A budget is the combination of intended expenditures along with proposals for financing them. However, because the goal is to support policy decisions, the budget process should be focused on policy decisions. One prominent recent budgetary process—the Gramm-Rudman-Hollings procedures—focused on the outcomes for spending and revenues, and was ultimately unsuccessful. In contrast, the subsequent focus on discretionary spending caps and PAYGO rules was focused on the legislative actions that Congress takes each year, a more desirable structure for a budget process.
Accurate and balanced in its presentation. Well-constructed budgetary presentations show accurately the fiscal implications of alternative choices. At the heart of such presentations is the ability to bring the lessons of sophisticated government cost-accounting to the projection of future outlays under a variety of scenarios. Similarly, the accumulated experience with alternative tax policies in the past should be the basis of constructing projections of future receipts.
Accurate projections, projections that ultimately encompass the direct budgetary impacts and the responses of the sub-federal governments and private sector into the estimates, have obvious value in presenting budgetary decisions. However, balance may be a more important attribute than accuracy—balance in the sense of treating all policy proposals in an identical fashion. Because the budget process must be comprehensive, it is essential that all choices be part of the analysis. In some circumstances, a novel proposal may present real difficulties in estimating the outlays or receipts. Choosing methods that can encompass all proposals—at the potential expense of accuracy in other areas—may ultimately serve the Congress better.
Tied to the financial management system. Ultimately, policy choices must be implemented by federal agencies. It is essential that the choices made by the Congress tie directly to the account structures, payment systems, procurement protocols and other aspects of the financial management system. In this way, the effectiveness of congressional control is enhanced. Moreover, when constructed in this fashion, accounting and audit information is directly amenable to analysis and improved estimates of budget proposals in the future.
Perceived Failures of the Current Budget Process
The current budget process is a source of broad dissatisfaction—some would even go so far as to declare it dead or broken. Placed in the context of the framework outlined above, a non-comprehensive list of complaints would include:
Lack of comprehensiveness. The current budget process does not yield comprehensive decision-making. Congress is widely perceived to undertake too little oversight of existing programs, missing the decision to terminate programs, eliminate duplication, or otherwise enhance the efficiency of operations.
Similarly, in recent years the increasing reliance on omnibus appropriations bills and frequency of supplemental appropriations legislation is rife with the potential for Congress to inadequately assess tradeoffs among alternatives—or fail to consider them at all.
And Congress may be missing the most important decision every year—the decision to allocate taxpayer resources to mandatory spending programs. In the years to come, the rising burden of mandatory spending is the central policy issue. At present, however, the budget process does not present Congress with any yearly decision regarding the scale of these programs.
Inaccurate and unbalanced presentations. The current budget process is rife with problems, beginning with the starting points: the congressional budgetary baseline prepared by the Congressional Budget Office (CBO) and the Administration’s baseline prepared by the Office of Management and Budget (OMB). The former is prepared under the assumption that current law will be unchanged over the budget window, while the President’s budget is presented under the assumption that all of the Administration’s proposals are implemented. This, in itself, is a continuous source of confusion.
In principle, it should be possible to prepare both budgetary presentations under common assumptions. However, over time it has become progressively more difficult to do so as the budgetary concepts and assumptions have shifted.
The result is a system in which confusion easily arises about the size of budgetary objectives and the degree of success in reaching them. The decisions made by Congress ultimately fall to the Administration to execute. Thus, the two processes should integrate easily, which is made difficult by the fact that they operate off different starting points.
The current budget process is also riddled with “labeling games.” Economically equivalent transfers to individuals are easily called either outlays or (refundable) tax credits. Flows of resources from the private sector to the government are labeled offsetting collections or offsetting receipts—negative spending—instead of revenues, even when there is no business-like transaction involved. Funds are labeled as “emergencies” even when widely anticipated, unrelated to the vagaries of natural disasters or war, or even responding to events in the relatively distant past.
The absence of effective enforcement procedures also leads to inaccurate presentations of tradeoffs. The fiction of an “advance appropriation” is meaningless in the absence of a genuine corresponding reduction the future of funds available to meet policy objectives. Similarly, the process of using fictional “mandatory spending reductions” in future years (which are quickly undone so as never to happen) to meet annual appropriations targets makes a mockery of a level budgetary playing field between either discretionary and mandatory spending or present and future outlays.
A final aspect of the current budget process is the increased use of sunsets and other techniques that take advantage of the 5-year budget window. Congress has long-employed sunsets to lower the apparent costs of provisions, so this aspect is not new. Instead, it has been employed with greater frequency in recent years. This has been most widely associated with tax legislation with the widespread assertion that Congress is mislead about the budgetary impacts. Given the attention paid, for example, to the 2003 tax legislation I find it difficult to believe that this is a large problem.
A more recent incarnation of this approach is somewhat more troubling. The budget resolution for FY2006 contained a point of order that the Budget Chairman could raise against mandatory spending legislation that raised outlays over the long term (5 separate 10-year windows were employed for the test). This had the potential to focus congressional attention on the central problem facing the U.S. budget, but instead it has generated a rising number of artificial sunsets to generate compliance with the provision.
The Way Forward
Even the relatively short list of problems above suggests that there is great potential to improve the budget process. The key question is how aggressive such an effort might be. One could imagine an approach that consists largely of band aids for the current process, or more extensive reconstructive surgery. To belabor the medical metaphor, one might also wish to contemplate wholesale transplant of a new process.
Band Aids. There are a host of provisions that have been suggested as possible improvements to the budget process: restoration of PAYGO rules, a rainy day fund and tighter enforcement of emergency designations, a line-item veto/enhanced rescission authority, a sunset commission, and earmark reform.
- PAYGO rules. Restoration of PAYGO rules is desirable. In some circles, PAYGO is associated with a reduced ability to undertake desirable tax policies. But this need not be the case. The greatest potential advantage to PAYGO rules is the ability to highlight spending disguised as tax credits and other “tax cuts” that have the economic impact of targeting resources toward particular activities. Moreover, the restoration of PAYGO has the potential to improve rather than impede pro-growth tax policies. At present, the U.S. income tax code is targeted on high-income individuals and corporations. For both groups it would be desirable to retain low marginal tax rates (indeed the high U.S. corporation income tax is a troubling feature of the tax landscape) and move the tax base to more closely resemble the consumed-income of those entities. PAYGO rules would place a premium on bringing consumption expenditures (for example, gold-plated employer sponsored health insurance for the highly-compensated) into the tax base as a desirable aspect of keeping taxes on the return to saving and investment low.
- Budgeting for emergencies is long-overdue. A straightforward approach is to mimic the mechanics of an insurance policy, an instrument that covers the financial consequences of fire, theft and other emergencies. Insurance policies charge a premium that is the “typical” or average level of losses. In some years, policyholders experience no losses and reserves build. In those years when losses do occur, reserves are drawn down. On average, the policy breaks even.
Congress could undertake a similar approach. First, it would require a tight and genuine definition of emergency. With this in place, the Congress could budget a fixed amount—say the average for the past ten years—for emergencies in each year, essentially charging itself the cost of an insurance policy against the particular emergencies in the definition. There would be no need for a “reserve” fund—the U.S. economy serves as the reservoir of resources. In those years when emergency spending was above average the result would be a larger deficit or lower surplus—requiring greater issuance of Treasury securities to transfer the resources from the reserve fund. In years when spending was below average, the reverse would be true.
- Line item veto/enhanced rescission authority. The Administration has proposed, and the Senate is considering, an enhanced rescission authority intended to mimic the authority of a line-item veto. To the extent that legislation is successful in granting an item-veto-like authority, on should not anticipate a dramatic impact on the level of outlays. The large body of research examining the impact of the (widely-varying) line item veto powers in the hands of state governors has two broad findings: (a) total outlays are little-affected, and (b) the composition of the outlays is affected by the ability of the governor to veto specific items. In short, the item veto is a powerful, but focused, weapon against the worst sorts of pork-barrel spending.
However, mimicking the line-item veto may not be the right objective. Rescission authority is in itself a desirable aspect of financial management. In light of the perceived failures of oversight, the Congress may wish to enhance the ability of the executive branch to conduct this activity. If so, it would be desirable to have the legislation permit adequate time to analyze recently-passed legislation for desirable rescissions—speed is less of a focus than getting it right—and fast-track procedures for the Congress to consider the proposed cuts.
- Sunset commission. Some observers have proposed a commission to examine the landscape of federal program and identify those that should be terminated. This is hardly an objectionable goal. However it raises more questions than it answers—what criteria will be used; how will such a commission be constituted; how will the Congress employ and deliberate on its findings; and so forth. More generally, the goal should be for the budget process to include sufficient oversight and review that such a commission is not needed.
- Earmark reform. The Congress is in the midst of earmark reform, and is to be applauded for this effort. The damage of earmarks lies less in the total number or dollars earmarked, and more in the damage to the integrity of the budget process. Budget should be an accurate, balanced, and comprehensive presentation of policy objectives. Earmarks are designed to remove balance and lessen the comprehensiveness of the tradeoffs.
Reconstructive Surgery. The Congress may wish to undertake more ambitious reforms of the budget process to more effectively address its current flaws. (In doing so, it does not preclude also adopting some of the less aggressive measures discussed earlier.) Among the candidates for such changes are creating a budget concepts commission, direct spending caps, stronger long-term points of order, a joint budget resolution, and biennial budgeting.
- A budget concepts commission. Since the passage of the 1974 Budget Act, which was in turn built upon the work of a 1967 budget commission, the array budgetary activities has expanded dramatically. As noted earlier, there has been a profusion of tax credits that have the economic character of mandatory spending provisions. In addition, over time there have been a rising number of accrual-like budgetary treatments—student loans, Pentagon health costs, etc. – present in the cash-flow federal budget. Further, the increasing scope of government financial activities—direct loans, loan guarantees, asset purchases and sales—raises the question of how best to present both the typical costs of such programs and the risks to which the taxpayer is exposed. The time is overdue for a group of technical experts to review the concepts on which the federal budget is based.
- Direct spending caps and long-term points of order. The most pressing fiscal policy for the federal government is the long-term path of mandatory spending. Budget mechanisms to accurately portray these costs and identify key decisions are desirable. Two approaches to doing so are to impose caps on direct spending and to institute stronger points of order against legislation with long-term budgetary impacts. In both instances, the key is to develop effective measures of the long-term consequences—no small feat. The key difficulty is how to treat legislation whose character clearly embodies long-term consequences, but which contains sunset provisions explicitly designed to eliminate the long-run consequences.
- A joint budget resolution. As mentioned earlier, a key difficulty in the current environment is the confusing array of budgetary baselines and proposals that emanate from the Administration and the Congress. A joint budget resolution would be a desirable means to bring the Administration and the Congress to a single set of measures. A second objective—to ensure timely and regular budgeting—is less clearly achieved. The joint budget resolution would carry all of the political difficulties of any piece of legislation that must pass both houses of Congress and be signed into law. Moreover, given its scope it likely carries more. While desirable, it would not be a panacea.
- Biennial budgeting. Proponents of biennial budgeting argue that an alternating cycle of legislating followed by oversight would improve the federal budget. Doubtless this is true, but it is far from obvious that this is what would happen. To my eye, the Congress needs to be involved more in budgeting, not less.
Budgetary Transplants. To some, the best approach to federal budgeting is to start over. In this approach, two different strategies stand out. The first is to transplant to the government the full array of private sector financial accounting—balance sheets, income statements, and cash flow statements—and budget on this basis.
This is a bad idea. Quite simply, the government is not like any private sector firm. Its two greatest assets are the sovereign powers to tax and to print money. What value should one ascribe to these assets? How do they change through time? The approach appears an awkward fit at the outset.
On the “liability” side of the ledger (“liability” put in quotes because there is an enormous gap between genuine legal liabilities like Treasury securities and the political promises embedded in federal programs), the largest numbers are associated with the future of health care programs. Under current law and using historic growth rates in health spending, federal outlays for Medicare and Medicaid will rise from just over 4 percent of Gross Domestic Product (GDP) at present to 22 percent of GDP in 2050.
Most observers assert that this “just can’t happen,” and they are correct. But what will happen? The Medicare Trustees assume that spending growth will ultimately slow by over 50 percent, but there is no evidence of such a decline thus far. In order to compute the appropriate balance sheet entry, it is necessary that arbitrary assumptions be made and these assumptions will dominate the outlook for the federal budget. Any approach that makes technical assumptions—not the legislated choices of the Congress—the focus is fundamentally flawed.
A second form of transplant is to adopt an entirely new budget system tailored to the needs of the federal government following the model of the Base Realignment and Closure Commissions. In this approach, a commission of independent experts designs such a system which would take effect unless disapproved by a joint resolution of the Congress. In this approach, the key transplant is not only a budget process, but the politics by which it is adopted.
Mr. Chairman and members of the committee, thank your for the opportunity to testify. Clearly there is great room for improvement in the federal budget process and I commend you for investigating the various approaches. I look forward to your questions.