FIGURE 7 Components of Net Federal Credit, Fiscal Year 1972 to Fiscal Year 1984 a. Primary guarantees, which exclude secondary guarantees and guaranteed loans acquired by onand off-budget agencies. b. Administration estimate. SOURCE: Office of Management and Budget, Budget of the U.S. Government, Fiscal Year 1984: 26-196 0-83--12 new types of loan-guarantee assistance programs have been initiated; these include the program to aid the commercialization of synthetic fuel production and the loan-guarantee assistance programs for New York City and the Chrysler Corporation. In part, the upsurge of federal credit activities in recent years was undoubtedly related to the increased demand for federal credit assistance as general credit conditions tightened. In part, however, it also seems to reflect the fact that while direct federal spending was being subjected to progressively tighter controls through the Congressional budget process, federal credit activities were, at least until recently, largely exempt from the discipline of that process. This has posed a major challenge for the budget process. THE CREDIT BUDGET The only credit activities included in the unified budget and subject to the normal budget process are on-budget direct loans and default payments under loan guarantees. However, even on-budget direct loans are not adequately controlled through the regular budget process. Most are shown only on a net-of-repayments basis. While this measures the net drain on credit resources in financial markets, it masks the total volume of new lending activity, which is often unrelated to repayments. Moreover, by far the larger share of direct lending is off-budget, consisting mainly of loans made by the off-budget Federal Financing Bank to borrowers who have obtained a guarantee from a federal agency. These and other new commitments for loan guarantees are excluded from the unified budget except in cases of default. To permit better overall control of federal credit activities, a partly overlapping credit budget was included in the President's budget starting in 1980. It provides estimates of all direct loans (both on- and off-budget) and loan guarantees, measuring these in terms of gross obligations (for direct loans) and gross new commitments (for guarantees). Congress has made increasing use of credit budget concepts and information in developing its budget resolutions. For example, the First Concurrent Budget Resolution for fiscal year 1983 included target ceilings for total credit budget direct loan obligations and loan-guarantee commitments. It also indicated how these totals should be allocated by budget functions. The Senate, moreover, made credit budget ceilings binding (that is, subject to a point of order if breached). The House voted for what was thought to be a comparable provision, but this was subsequently declared invalid on technical grounds. The recent procedures to bring federal credit activities under better control have been developed largely on an experimental, ad hoc basis. As a next step, we believe that credit programs should be permanently subject to most of the budget-control procedures now used in connection with direct spending. In particular, there should be a formal requirement that budget-resolution ceilings on total new loan extensions and loan-guarantee commitments be made binding. Because credit activities subject to this control process would be those included under the credit budget, the proposed ceilings would apply automatically to both off-budget and onbudget items. These ceilings would essentially be directed at limiting the impact of federal lending and guarantee programs on private credit markets. Although the recommended procedure is clearly desirable as part of an overall effort to bring credit activities under better budgetary control, it would pose a number of problems if it were not followed by additional steps. One problem is to determine what the appropriate limits on credit extensions should be. Because different federal credit programs have widely varying impacts on private financial markets, it cannot be automatically assumed that particular federal credit programs will necessarily crowd out private credit activities on a one-to-one basis. For example, if Federal Housing Administration (FHA) loans were not available, private borrowers would to a considerable extent still turn to other lenders. Moreover, the degree of crowding out tends to vary with the overall strength of credit demands at different periods of time. Hence, considerable further study of the influence of federal credit activities on financial markets and submarkets is needed to assure that ceilings on such activities are set at levels that yield optimum economic results." INTEGRATING CREDIT AND DIRECT SPENDING DECISIONS An even more serious problem stems from the fact that most of the transactions included in the credit budget are not part of the unified budget. Hence, budget control procedures based on the credit budget are not integrated with the regular Congressional budget process, which covers only unified budget transactions. There have, in fact, been some suggestions that it would make sense to separate the unified budget and the credit budget completely, mainly by taking on-budget direct loans (which are already counted in the credit budget) out of the unified budget. This would 6. Questions can also be raised regarding the range of activities to which budget control mechanisms directed at federal influences on credit markets should extend. Should federal support for secondary markets for loan instruments be included? Even more broadly, it could be argued that loan guarantees are merely a special case of a government guarantee of the performance of one or more parties in a private contract. Should all such guarantees be included in the budget, and if so, in what way? Questions of this type should be carefully examined by the proposed Budget Concepts Commission. eliminate the existing overlap and provide two separate budgets for control purposes, one for spending and the other for lending and guarantees. Under this arrangement, the credit budget total could then (with some adjustments) be added to deficits from spending and taxing operations to produce a measure of total federally induced drains on credit markets and thereby provide a basis for limiting such drains. But such a dual-budget system would also have a major disadvantage: it would, in practice, prevent direct comparisons and trade-offs between spending and credit programs having the same or similar purposes. This is a major shortcoming to the extent that there is concern with the budget's effects on incomes as well as on financial flows and with its role as an allocator of governmentally provided resources among competing uses. With separate spending and credit budgets, for example, there would be no way to make a direct comparison between an outright federal grant for education and a federal loan program for the same purpose. Yet, some of these loan programs are so highly subsidized that they are virtually equivalent to outright grants. We recommend that in addition to relying on a credit budget, the Administration and Congress make strong efforts to develop an integrated approach to budget control that will, as far as possible, allow direct tradeoffs between credit and spending programs within the unified budget. A major obstacle to full integration of credit and spending programs is that these programs cannot be readily compared on a dollar-for-dollar basis. Of course, no one is suggesting that there should be a direct trade-off between a government expenditure, such as acquisition of a tank or a grant. to a locality for school lunches, and a loan or guarantee program of an equal dollar amount. What is really needed is a comparison of the subsidy elements of loan or guarantee programs (including a risk premium) and direct spending. How can this be done? One procedure that has been suggested would begin by weighting each dollar of federal credit (loan or guarantee) according to the discounted present value of the subsidy involved. (The President's fiscal year 1984 budget presents this type of information in detail.”) Actual loss experience under particular credit programs could also be taken into account in estimating the subsidy value. Each weighted amount would then be used as a guide to decisions on allocating budget resources to alternative uses. For example, when a spending committee is asked to reduce spending for programs under its jurisdiction by a specified amount, the budget committee might give it the option of achieving part of the cut by reducing 7. Office of Management and Budget, Budget of the U.S. Covernment, Fiscal Year 1984, Special Analyses (Washington, D.C.: Government Printing Office, 1983), F-54-60. the subsidy element of a loan program rather than by basing the cut entirely on a reduction in direct spending. As initial steps to help integrate federal credit activities into the regular budget process, we recommend, first, that further emphasis be placed on measuring the interest-subsidy elements in federal loans and guarantees and disseminating such information in readily understandable form and second, that there be experimentation with budget procedures that involve direct comparisons between regular expenditures and the subsidy elements of credit programs. Even these procedures, however, would not fully integrate credit programs into the existing trade-off mechanism within the unified budget; outlay equivalents of credit and guarantee programs based on the present value of the subsidy would still not be formally counted as part of the unified budget. And unless the items to be traded off are automatically part of the same total, a trade-off procedure may be hard to enforce. The question thus arises whether a more straightforward approach could be developed under which the subsidy elements of credit programs would be explicitly incorporated in the unified budget and, consequently, in the regular budget control process. A major problem with such an approach is that determination of the present value of a credit subsidy involves complex calculations and leaves room for considerable differences of opinion on what constitutes the best measure. Formal inclusion of computed credit-subsidy figures in the unified budget might thus create irresistible temptations to play games with budget numbers. This would run directly counter to the principle that the budget should be readily understandable and subject to as little manipulation as possible. In theory, this difficulty could be overcome if the federal government were to stop making direct loans and guarantees altogether, if it provided those in need of subsidized credit with cash grants equal to the present value of the subsidy, and if it then left it up to the would-be borrowers to use these subsidy funds to obtain the needed credit on the desired terms directly from financial institutions. In actuality, of course, such an arrangement would probably not be feasible in most cases. We believe, however, that a variation of this approach might prove practical and deserves careful exploration. It would involve the establishment of a national lending fund that would carry out all federal loan and guarantee activities but would not be allowed to subsidize transactions or take risks on its own. All government agencies requiring guarantee or direct lending programs for their clientele would have to pay sufficient sums to this fund to cover the subsidy elements of the credit programs plus a reserve |