more than 8 million Americans out of work. Nevertheless, inflation was extremely high, slightly in excess of 9 percent as measured by the Friedman formula. Had this spending limitation been in effect, it would have constrained budget outlays for fiscal 1977, the first full fiscal year after calendar year 1975. Total outlays in fiscal 1977 would not have been allowed to exceed $395.4 billion, $10 billion below the amount (including the off-budget entities) initially proposed by President Ford in what was widely regarded as an extremely tight budget, and $17 billion below the final results for that year. Without taking account of automatic adjustments in the economy (for example, a smaller budget might have led to lower GNP than was actually produced), federal outlays would have been about 21.5 percent of GNP, a drop of more than 11⁄2 percent compared to the preceding fiscal year. Thus, in lieu of a gentle decline in GNP promised by the Friedmen group, there might have been a steep, destabilizing drop during a period of continuing high unemployment.

The inflation penalty is intended to prod Congress to do something about inflation. In the words of its sponsors, "the higher the inflation rate, the greater the incentive for Congress to reduce inflation." The assumption is that inflation is always within the control of Congress, a position that ignores other contrib uting factors such as world markets and the actions of other countries.

At the core, the three spending limitations reviewed here seek a forced retrenchment in the size of the federal government. Their principal aim is to assure less rather than more government. McClure would require a modest cutback that probably could be accommodated within conventional budget methods. The cuts might be painful, but they could be achieved without requiring a significant redirection in the scale and priorities of the federal government. Dole would compel a much steeper reduction, amounting to about one-sixth of the size of the federal budget. Unless it was phased in over a number of years, the Dole amendment would be an abrupt turnabout in the role and scope of the federal government. Of course, Dole's concept could be retained but its impact mitigated by pegging the expenditure limit at above 18 percent of GNP. If 20 percent were set as the constitutional limit, it probably could be achieved without a disruptive effect on public policy.

The Friedman approach is open-ended with its impact depending on future economic growth and inflation. But it is possible that over time this could have a much more radical effect on the federal government than either of the other spend ing limitations. Unless inflation is significantly lower than it has been during the past decade, the size of the federal government will have to contract in order to abide the Friedman limits. Although the reductions would be decremental rather than all-at-once, the long-term consequences could be truly radical. One wonders whether a fundamental reordering of the role and purposes of the federal government should be considered under the guise of a spending limitation. The issues raised by Friedman are far more than budgetary matters alone, for they reach to almost every corner of public policy.

OVERRIDING THE CONSTITUTIONAL RESTRICTIONS

Most of the proposed constitutional amendments would authorize Congress to temporarily set aside the restrictions. While the procedures differ, they usually provide for a concurrent resolution (a legislative measure which is not presented to the President for his review) specifying that conditions warrant deficit spending or expenditures in excess of the constitutional limitation. A few (such as Friedman) predicate congressional override on a declaration of emergency by the President, but most enable Congress to act unilaterally without the participation (or concurrence) of the executive branch.

The proposed amendments differ as to the conditions that would permit an override.

Some refer to "national emergency," others to "grave national emergency"; some mention war, other speak in more general terms. As long as Congress can decide whether a particular condition meets the requirement, the differences in wording are of little consequence. However, the size of the congressional majority required to override could make a big difference. The requirement ranges from a majority of the voting members of the House and Senate in S.J. Res. 11 (Talmadge) to three-quarters of all the membres of the House and Senate in S.J. Res. 7 (Armstrong). In a close vote, the numerical standard could spell the difference between victory and defeat for an override attempt. It bears noting that in recent years most of the concurrent resolutions on the budget have passed the

House with razor-thin majorities, sometimes fewer than a handful of votes. With House Republicans lined up in virtuallly solid opposition to deficit spending, the resolutions have barely managed to squeak through. A high constitutional threshhold for overriding the limitation is likely to bolster the bargaining positions of congressmen who favor smaller deficits or lower spending. By withholding their support until the deficit or spending level is brought closer to their preference, the holdouts would be able to magnify their influence over the budget. Thus, while a high requirement would not necessarily rule out imbalanced budgets or breaches of the spending limitations, it might lead to smaller deficits and less spending by Congress.

Some of the proposed amendments would place substantive restrictions on the power of Congress to override a balanced budget requirement. S.J. Res. 16 (Wallop) would not permit the deficit to exceed total revenues by more than 10 percent. At current budget levels, this would allow a deficit of about $50 billion. S.J. Res. 5 (Dole ) would not perniit a deficit in more than three years out of any eight consecutive years. If this restriction were to have retroactive effect, it would bar deficit spending during the first 5 years that the constitutional provision was operative. Repaying the Deficit and the Public Debt

A number of the proposed amendments provide for Congress to liquidate any deficit incurred after the constitutional restrictions take effect. S.J. Res. 2 (DeConcini) and S.J. Res 6 (Stennis) require that any deficit is to be paid off by a surtax levied in the next calendar year. The surtax would become effective without congressional action; all that would be required is a presidential message setting the surcharge at a rate sufficient to cover the deficit. However, Congress could suspend all or part of the surtax by a two-thirds (S.J. Res. 2) or threequarters (S.J. Res. 6) vote. S.J. Res. 5 and S.J. Res. 16 allow more time for repayment, up to three years in the former and as much as four years in the latter. A few of the proposed amendments would require the accumulation of surpluses to retire the public debt. S.J. Res. 7 would spread repayment of the debt over 100 years, with 10 percent repaid during each 10 year period. S.J. Res. 16 would establish a 20-year repayment schedule, with annual surpluses equal to 5.5 percent of total outlays used for this purpose.

What's in a Budget?

Any constitutional restriction on deficits or expenditures must come to grips with the issue of what constitutes the budget of the United States Government. The question of what ought to go into the budget is by no means settled; current practices are a compound of written and unwritten rules, many of which were introduced by the executive branch without the explicit concurrence of Congress. Moreover, the emplacement of an expenditure or budgetary restriction in the Constitution will not bring an end to the improvisation of new governmental forms and accounting procedures. Quite the opposite, one can expect that the more formidable the restriction the more likely that it will invite the executive and legislative branches to contrive new practices. Only one thing can be certain: the more inclusive the definitive of the budget, the more restrictive will be any constitutional limitation on deficits or expenditures.

Off-budget programs.-The leading candidates for inclusion in the budget are six federal agencies all or portions of whose activities have been excluded from the budget's totals by Congress. The six off-budget agencies are the Federal Financing Bank (FFB), the Rural Electrification and Telephone Revolving Fund, the Rural Telephone Bank, the Pension Benefit Guaranty Corporation, the Postal Service Fund, and the U.S. Railway Association. A seventh, the Board of Governors of the Federal Reserve System, has a somewhat ambiguous status. It is excluded from the budget by tradition rather than by law.

Off-budget outlays are expected to total $12 billion in fiscal 1980. More than $11 billion of this amount is accounted for by the purchase of federally-guaranteed obligations by FFB. Under its charter, the Federal Financing Bank is authorized to acquire any obligation guaranteed in whole or in part by any federal agency. (It also can purchase direct federal obligations, but these are included in the budgets of agencies making the loans.) There is not limit on the amount of guaranteed loans that can be purchased by FFB. When it needs funds to finance these purchases, FFB borrows from the Treasury which, in turn, provides for its cash needs by borrowing from the public. Thus, FFB's loan activities have the same impact as the regular budget deficit on capital markets.

The House Budget Committee recommended in 1976 (H. Rept. No. 94-1740) that various off-budget agencies be included in the federal budget, but it withheld a recommendation on the status of the Federal Financing Bank pending further study. However, in 1978 (H. Rept. 95-1055, p. 23) the Budget Committee supported legislation to include FFB in the budget. If all off-budget agencies were placed on-budget, there would be an estimated net addition of $12 billion to the fiscal 1980 outlays and deficit. Presumably, this additional amount would be covered by a budget or expenditure limitation written into the Constitution. Government-sponsored enterprises.-These enterprises are privately-owned corporations established and chartered by the United States to perform certain "public" functions. Seven such enterprises are identified in the fiscal 1980 budget, and their financial statements are annexed to the Budget Appendix. However, the finances of these organizations are not included in the outlay or deficit totals of the United States Government. The seven sponsored enterprises are credit operations: Student Loan Marketing Association; Federal National Mortgage Association; Banks for Cooperatives; Federal Intermediate Credit Banks; Federal Land Banks; Federal Home Loan Banks; and the Federal Home Loan Mortgage Corporation.

These enterprises are only a few of the hundreds of corporations created or chartered by Congress since 1791. During the past two decades, Congress has devised a variety of organizations forms combining elements of public and private enterprise. These include private corporations which are funded entirely by federal appropriations (e.g., the Legal Services Corporation) private for-profit corporations which have public and private sources of revenues (Consolidated Rail Corporation), and profit-making corporations partly owned by the federal government (National Railroad Passenger Corporation).

The 1967 President's Commission on Budget Concepts devised a simple test for determining whether a government-sponsored corporation should be on- or off-budget:

"The criterion recommended by the Commission is basically that Government sponsored enterprises be omitted from the budget when such enterprises are completely privately owned."

As the activities and finances of the "quasi-governments" (as they are sometimes referred to) have become more complex, it has become increasingly difficult to use the private ownership test to distinguish between public and private activities. In a 1978 paper prepared for the National Academy of Public Administration, Graham S. Finney argued that "an either/or, public/private world, separating government and for profit enterprises, has long ceased to exist, if it ever really did." But whether or not such distinctions are valid, they would have to be made in applying a constitutional limitation on the budget. Government sponsored corporations would have to be placed either within or outside the budget; no gray areas could be tolerated in deciding whether or not a quasipublic expenditure creates a deficit or exceeds a spending limitation.

The private ownership test is not a full measure of the federal government's involvement-financial and administrative-in the activities of sponsored organizations. With respect to the seven privately-owned enterprises annexed to-but no included in-the budget, the federal government has a considerable interest in their finances and performance. Their financial relationship to the federal government is described in the 1980 budget (Special Analysis, p. 164): "Since they are private, their activities are not included in the budget totals. They are all subject, however, to some form of Federal supervision and by law or by custom consult with the Treasury Department in planning the marketing of their debt. . . .

"Government sponsorship has endowed these enterprises with certain characteristics that differentiate them in credit markets from completely private institutions. They have been given special preferences, and certain tax exemptions and the special eligibility of offering their securities as investments of federally regu lated institutions. These advantages give their security obligations a preferred position in the securities market, enabling them to borrow at rates only slightly higher than those of the Treasury."

Although the securities issued by these corporations disclaim any liability of the United States, they are similar to the "moral obligation" (or revenue) bonds issued by public authorities in many states. They are listed in the New York Times and other publications under the heading, "Government and Agency Bonds." Despite the legal disclaimer, the federal government probably would not be able to remain on the sidelines if any of the enterprises fail. The involvement of the United States is potentially greater than that of stockholders of

private corporations whose exposure is limited to the value of their stock. The private ownership test might be an irrelevant standard by which to measure the public character of federally-sponsored enterprises. In its 1976 report on off-budget financing, the House Budget Committee explicitly rejected private ownership as the sole criterion for establishing the budgetary status of these enterprises. It noted that in most cases investment markets do not consider

"the Government-sponsored enterprises to be completely private nor, in most cases, would a member of the general public draw this distinction. Each of the Government-sponsored enterprises is subject to some greater or lesser measure of Federal direction, and some implicit subsidy may be presumed to arise whenever borrowing from the public is undertaken." (H. Rept. No. 94-1740, p. 7)

In accord with this finding, the Budget Committee recommended that Congress not create any new enterprises with off-budget status "pending resolution of the question of appropriate criteria for existing agencies."

The sponsored enterprises will account for an estimated $19 billion in outlays in fiscal 1980. It should be noted, however, that neither Congress nor the President exercises direct budgetary control over these loan activities, so that actual outlays can vary significantly from the original estimate.

One can anticipate that stringent constitutional restrictions on deficits or expenditures would invite the creation of additional "quasi-governmental" enterprises outside the budget. This clearly has been the experience in state governments, a number of which have turned to public corporations and "moral obligation" bonds in order to evade their own constitutional restrictions. States have financed billions of dollars of capital construction by vesting responsibility in private or quasi-public organizations which are not legally deemed to be governmental entities. These evasive practices often are more costly than direct borrowing by the state because revenue bonds generally bear higher interest rates than comparable general obligation bonds. Furthermore, these special institutions often escape not only constitutional limits on borrowing but also the political controls to which regular public organizations are subject.

It would not be difficult to remove tens of billions of dollars from the federal budget by creating "quasi" institutions with authority to raise revenues and to borrow. It would be possible, for example, to create a $100 billion dollar national health insurance system without a single dollar for this purpose appearing in the budget. While there may be legitimate reasons for creating these types of enterprises, overly restrictive constitutional controls can have a warping effect on the structure of the United States Government.

Guaranteed loans.-The third major category of federal financial transactions excluded from the budget total consists of loans for which the United States commits itself to pay all or part of the principal and or interest in case of default by the borrower. Guaranteed loans are excluded from the budget by section 3(a) (2) of the Congressional Budget Act which provides that the term budget authority "does not include authority to insure or guarantee the repayment of indebtedness incurred by another person or government." This exclusion is based on the fact that guaranteed loans (unlike direct loans which are in the budget) are a contingent liability in which the obligation of the United States is activated only in case of default. The United States bears no liability if the borrower makes the required debt payments; in case of default, any payment by the United States is included in the Budget.

The budgetary status of guaranteed loans was considered by the President's Commission on Budget Concepts in 1967. In comments prepared for the Commission, Dr. Arthur Okun predicted that the different budgetary treatment of direct and guaranteed loans "would lead to a strong preference in the budgetary process for guarantees over direct loans." Moreover, the staff of the Commission cautioned that

"***continuing pressures on both Congress and the Executive to hold down the level of apparent Federal spending and the budget deficit establish artificial incentives to shift from direct loans to indirect lending whether or not the guarantee or insurance of private loans represents a fully satisfactory substitute." These expectations have been realized. There has been a huge increase in the volume of loan guarantees. By the end of fiscal 1980, the total of such loans outstanding is projected to exceed $400 billion, an increase of almost $100 billion in the space of just three years. The total of such loans held by the Federal Financing Bank is expected to reach $56 billion, and the amount held by the public is estimated at $239 billion.

Various congressional committees have expressed alarm at the growth in these guarantees. In a 1976 report, the House Appropriations Committee was disturbed by

*** the recent proliferation of legislative proposals involving program guarantees, which while they do not grant budget authority in the technical sense of the term, do thus establish a contingent liability on the taxpayer virtually beyond the subsequent control of Congress."

In a report on the second budget resolution for fiscal year 1977, the Senate Budget Committee voiced concern about "the increasing use of guarantee mechanisms that have the effect of circumventing the discipline of the Congressional budget process," and it indicated its desire for "appropriate recommendation to Congress to establish control over proliferation of such programs.'

The 1980 budget takes two steps toward budgetary control cf loan guarantees. For the first time, schedules of loan guarantees have been included in the Appendix for each account with which such guarantees are associated. In addition, the Administration has proposed that annual limits be set on outstanding loan guarantees. Even if such limits were established, one can anticipate that constitutional restrictions on direct budget outlays will generate pressure for circumvention by means of guarantees.

The three off-budget devices discussed here will net no more than $50 billion in fiscal 1980. Although they presently are the chief means for conducting financial activities outside the budget, new mechanisms might be devised in response to constitutional restraints.

What is an Outlay?

If definition of the budget's scope is difficult, determining what constitutes an outlay can be even more troublesome. There is almost no law on the subject, and a jumble of conflicting practices and traditions complicates the quest for defensible criteria. Yet a spending limitation cannot be enforced without rules for deciding what is an outlay. The issue, at first glance, seems straightforward and without complication: an outlay occurs whenever the federal government makes a payment of funds. The issue, however, is not whether a payment constitutes an outlay, but how outlays are computed for budgetary purposes. It is in the counting of outlays that difficulties abound.

Grossing versus netting. In the course of each year, the federal government makes a sizeable number of direct loans (which, as noted earlier, are reported in the budget) and also receives payment from borrowers to whom loans were extended in previous years. The federal practice is to count direct loans in the budget on a net basis, that is the gross amount of new loans minus repayments. This method is consistent with the definition of outlays in the Congressional Budget Act as "expenditures and net lending of funds."

The financial activities of public enterprise revolving funds also are reported in the budget on a net basis. These funds conduct business-type operations for which they obtain receipts from the users of their services. The receipts, however, are not computed in the budget as federal revenues, but offset the expenditures of the public enterprises. As a result of this accounting practice, the amount shown in the budget as outlays for public enterprise funds is the excess of their expenditures over revenues. In fiscal 1980, the gross outlays of the public enterprise funds will be about $46 billion, but with $38 billion in receipts from the public, the budget will show only approximately $8 billion in outlays. The treatment of these outlays on a net basis conforms to the 1967 recommendation of the President's Commission on Budget Concepts.

However, in 1977, the Comptroller General recommended that these funds be reported on a gross rather than a net basis. The reasons set forth in his report relate to the ability of Congress to control the budget.

"*** when new accounts are classified [public enterprise] revolving funds or whenever existing revolving funds grow, budget outlays do not increase as rapidly as they would if accounted for on a gross basis. The changing of an account to a revolving fund gives the appearance that budget expenditures are declining even though the rate of expenditures may remain the same or even increase. Using the revolving fund classification might help to reduce budget totals and escape national priority debate by changing fund structure without changing fund purposes."

In other words, changes in accounting procedures and budgetary definitions can veil the actual budget totals. However, no constitutional restriction can be immunized against one or another form of budgetary legerdemain.