Skip Navigation

Mr. Chairman and members of the committee. I am pleased to discuss with you again the need for systematic congressional review of ongoing governmental programs and the strengthening of the congressional authorization process. A year ago, I testified on some of the general issues raised by S. 2925, the Government Economy and Spending Reform Act. S. 2, the Sunset Act of 1977, the successor to S. 2925, embodies a number of improvements from the earlier proposal. Mr. Chairman, you and members of the committee have made significant progress in your deliberations 6n this measure. In my view, this bill represents an important step beyond the Congressional Budget Act in the continuing effort to improve congressional decisionmaking processes. It deserves a thorough review by the Congress.

Needed Improvements in the Budget Process

The new congressional budget process has had a very significant and positive impact on congressional decisionmaking. The Congressional Budget Act of 1974 provided the Congress with a set of procedures that enabled it to deal with the budget as a whole and to focus on fiscal policy and other key issues affecting the budget totals. However, the expected improvements in the Congress' ability to debate national priorities and decide on the relative emphasis among major governmental activities--as envisioned by the drafters of the Budget Act--have not materialized, I believe, for two reasons.

First, the Congress does not have a process for systematically reviewing the existing programs and activities of the federal government to determine if they are accomplishing their objectives and that their benefits exceed their costs. S. 2 is intended to provide the Congress with such a mechanism and would provide a greater role for the authorizing committees in the congressional budget process.

The Congress already reviews periodically the operation of many government programs. Annual authorization measures are adopted for defense procurement and construction, foreign aid,and various research and development activities. Most grants to state and local governments have limited authorizations that are renewed every three to five years. However, this periodic congressional review is not done for a large number of federal activities where this would be appropriate, such as various tax expenditures, income transfer programs and regulatory activities. Also, congressional review is not carried out in a coordinated manner so that all activities serving a common function or similar objectives are reviewed at the same time.

Moreover, there is no systematic process for collecting and presenting to the Congress program evaluation information that could be used to judge the relative effectiveness of government programs and activities. Mr. Chairman, S. 2 addresses these problems and would make significant improvements to the congressional review of program performance and accomplishments.

A second reason why the Congress has not been able to debate and decide questions of national priorities in a satisfactory manner is that it still deals with the federal budget one year at a time. I believe it has become apparent to Members of Congress with the first two years of experience under the Budget Act that there is very little the Congress can do to make the federal budget next year look much different from that of this year, especially in terms of relative budget shares. A significant portion of each year's spending reflects legal commitments from prior years, for example, payments now due on an aircraft carrier ordered several years ago. An even larger portion is mandated by existing laws, such as social security benefits, interest on the public debt, and reimbursements to states for medicaid and public assistance expenses. More than 70 percent of federal spending is determined by advance spending decisions of one kind or another. Within the remaining 30 percent, it is very difficult--and often undesirable--to make sharp changes from the previous year's budget.

However, while changes in the budget in any one year are usually marginal, such changes may imply a major reordering of future priorities. Thus, a new weapons system, a major energy resource development or a new entitlement program may have costs that are small in the first year, but that grow to billions of dollars in future years. Similarly, providing a new tax preference may make little difference in the first year but produce substantial losses in future tax revenues.

The Congress already makes a great many budget decisions on the basis of future plans, predictions, and estimates. But it does not do so consistently, and it completes its decisions every year without ever adopting a description of their likely overall impact beyond the next fiscal year. Since it has not considered that outyear impact on the budget as a whole, the Congress has made no articulated judgment about whether it likes the direction in which the budget is headed or that it will reflect the right set of priorities. If the goal is a balanced budget or a 4 percent unemployment rate or both or something else in 1981, it is almost a certainty we will not get there if we wait until 1980 to state the goal and do something about it. The Congress should begin now to state these goals and to shape its annual budgets accordingly.

Advance Targeting

On February 24, I submitted a report to the Congress pursuant to section 502(c) of the Congressional Budget Act that discusses the need for planning the federal budget over a longer period than just one year ahead, The report concludes that it would be both advisable and feasible for the Congress to set budget targets farther in the future, a process we called "advance targeting." While not committing the Congress to any specific taxing or spending decisions, advance targeting would provide a context for making these decisions with their implications for outyear goals clearly in mind.

The Congressional Budget Act of 1974 already provides the procedural framework for advance targeting. All that is required is a decision to use it for that purpose. Our report outlines a possible implementation schedule that would permit the Congress to move into an advance targeting process in a series of stages over a three-year period. The major recommendation in the report is that the two Budget Committees formulate and the Congress adopt a plan for stating and voting on advance budget targets, with the eventual goal of annually adopting targets not only for the budget year but also for the four following years.

We recommend that the Congress start next year by setting targets for two years--fiscal years 1979 and 1980. The Administration should be requested to submit budget proposals for both years, as President Ford did in his last budget submission. Similarly, the March 15 reports from Congressional committees submitting their views and budget recommendations should cover two years.

Advance targeting is a method of focusing on the future implications of budget choices. The method itself is neutral. It will neither ensure nor prevent the arrival of the elusive fiscal dividend. But it will give the Congress the opportunity to decide that question, which it cannot do now because it does not adopt any outyear goal against which this year's actions can be measured. The targets to be stated, the array of choices to be considered, and the making of choices are, as we all understand, what the political process is all about. I believe that the Congress should adopt advance targeting as one method with a high potential for more informed choice and more satisfactory outcomes.

Mr. Chairman, with your permission, I would like to submit for the record our report on advance budgeting. It is a brief report and somewhat unique for CBO in that the law required us to make recommendations.

The Sunset Act of 1977

Mr. Chairman, I would now like to comment on some of the specific provisions of S. 2, the Sunset Act of 1977. First, I wish to commend you and other sponsors of the bill for the improvements that have been made in the bill from the original version on which I testified a year ago. The application of the sunset concept to the legislative authorization process requires careful thought and discussion. Where the Budget Act reforms were built upon a long established process, S. 2 would establish in many instances a whole new process. In this sense, S. 2 would be an even greater challenge to the Congress than the implementation of the Budget Act.

Two changes which I believe are especially important in S. 2 are the inclusion of tax expenditures in the periodic congressional review process and the adoption of less rigid procedures for the committee reviews. The effect of tax expenditures can be as great as or greater than the effects of direct government spending in achieving various social and economic goals and objectives. The analysis of the performance of various federal programs and activities would not be complete without including tax expenditures.

The adoption of a sunset review in place of a zero-base review also represents an improvement in the bill. This should give the authorizing committees more flexibility, and distinguishes the legislative review of authorizations from the executive branch use of zero-base budgeting concepts.

While S. 2 is a marked improvement over S, 2925, I believe some problems remain that the committee should give attention to. One problem involves the potential workload for the authorizing committees. The committee has already given much attention to this problem, and has taken steps to deal with it, including extending the review cycle from four to five years. Nevertheless, the magnitude of the work involved remains an unknown, largely because the number of programs and tax expenditures are uncertain and because the Congress has little or no experience in conducting 3 thorough and systematic review of existing programs and activities with similar goals and objectives.

S. 2 recognizes that the first step in moving to a more systematic congressional review process is the preparation of an inventory of existing spending programs and tax expenditures. The definition of programs involves many problems, which probably should be worked out before specific review dates are adopted. A primary problem is level of detail. One person's program is another person's program component. To develop a comprehensive inventory of existing programs and activities within the next nine months (by January 1, 1978) as required by section 102 that will prove satisfactory to all the committees involved will be a difficult task. I realize that a great deal of work to develop such an inventory has been done already by the General Accounting Office, but I believe that more work will be needed as the inventory is put into the context of a five-year review cycle.

A second potential problem involves the time schedule for the reviews. Under Title II, the reviews are scheduled to begin on or before October 1 of the year preceding the year in which the program authorization expires. Committee reports on the sunset reviews are due six months later--on or before May 15 of the year of the authorization expiration, which would occur on September 30 if the Congress took no action to extend the program.

As part of our study of the advisability and feasibility of advance budgeting, we conducted a special examination of the impact of federal budget decisionmaking on the operation of state and local government programs that are financed by federal funds. One of the things we learned in this study is that the planning needs of state and local legislative bodies and program administrations require early congressional decisions on funding levels for federal grants to state and local governments. One of the greatest complaints of state and local officials is that most federal grant programs involve uncertainties in funding, with a consequent sacrifice of program effectiveness. This uncertainty stems in part from the fact that at every level the participants are on different cycles. They also participate sequentially, and so even when the cycle variances are overcome, delays along the line bear most heavily on the end--the program operators and the people they serve.

The dilemma is that the longer the Congress waits for the most complete information on which to base its decisions, and the longer the federal administrators take to draft regulations and instructions to applicants, the less time the state or local administrator has to design a program, hire employees, locate the people to be served, and begin to serve them before the cycle is over and the uncertainty starts again.

Based on these findings, and the views of an advisory panel of state and local officials who participated in designing and reviewing our study, we made two recommendations in our advance budgeting report:

1. That the Appropriations Committees evaluate present federal programs for funding state and local activities, using criteria stated in the report, to determine whether any programs not now funded a year or more in advance should be so funded.
   
   
2. That the Congress (a) consider a rule for completing reauthorization action on federal-state-local programs a year before expiration of the authorization, and (b) extend the practice of fixing firm statutory deadlines for federal agency action on submitted state and local applications.

The second recommendation has particular relevance for the time schedule for sunset reviews in S. 2. I recommend that the committee amend the bill so as to have these reviews completed at least a year before the expiration of a program authorization. This would avoid creating unnecessary uncertainties for program administrators, which can have deleterious effects on program performance. It would also provide for a more orderly phaseout of programs that are to be terminated.

A third potential problem involves the substance of the sunset reviews. As I stated in my testimony last year, it is often difficult to measure the effect or performance of government programs. Program goals may be general, numerous, or difficult to quantify. There may be no agreement on appropriate evaluation criteria. The results of evaluation studies are often inconclusive. The full effects of terminating an activity may be impossible to determine. Thus, the objective of S. 2 to require a more thorough review of existing programs and activities and to weed out the ineffective ones will require painstaking effort over several years to improve evaluation methodology and data collection.

Unfortunately, S. 2 does not provide any built-in mechanism to ensure that the committee reviews are more than perfunctory. The sunset provisions are meant to provide this assurance, but they may have little effect. If there is strong political support for a program, it probably will be extended no matter what the results are from any evaluation studies. The authorizing committees, like federal agencies, become attached to their programs and generally are advocates for their continuation. S. 2 does not provide for any internal watchdog to monitor the review process and challenge the authorizing committees reviews if they are not thorough or convincing. The Congressional Budget Act established the two budget committees to monitor and guide the new congressional budget process. Perhaps the Budget Committees or some other pair of committees should be designated to perform a similar role for the sunset review process.

A Possible Implementation Approach

Last year I suggested that the implementation of the sunset concept might best proceed in a series of stages. I still believe this to be appropriate because of the many unknowns that are involved in undertaking this major reform. I will be more specific this time and suggest a possible series of stages for the committee's consideration.

First, I believe it would be useful to separate the provisions for the review of new programs from those of existing programs. A first step would be to establish immediately a requirement that all new legislative spending authorizations and tax expenditures have time limits not exceeding five years or have specific review dates where an authorization time limit is not judged to be appropriate. In addition, the sunset review procedures could be applied to programs coming up for reauthorization in 1978 and 1979. At the same time, work could proceed on the inventory of all existing programs, authorizations and tax expenditures. Following the preparation of this inventory, the views of the executive branch, state and local governments, and the public should be obtained on the need to terminate, consolidate, or otherwise change existing programs to improve program effectiveness and efficiency.

The inventory of existing programs and the recommendations of the executive branch and others would provide a solid basis for working out a five-year review cycle and provide the committees with a set of specific problems to address. Meanwhile, the application of sunset review procedures to new programs and those whose present authorization expires during the next two or three years would provide valuable experience in conducting these reviews.

This evolutionary approach to implementing the sunset concept should avoid many of the problems that otherwise could occur with an across-the-board implementation that the Budget Act procedures benefited from a similar evolutionary implementation.

In conclusion, Mr. Chairman, I want to reiterate my support for the periodic and systematic review concept. I believe it offers a major contribution to the Congress' decisionmaking procedures and would meet one of the major problems still remaining with the congressional budget process. The Congressional Budget Office will provide whatever cooperation and assistance we can to help make this important reform work. Thank you for this opportunity to discuss this proposal with you and I will be pleased to answer any questions you or members of the committee may have.