CRS REPORTS THAT PRIVATE-ACTIVITY BONDS HAVE FLOURISHED, DESPITE EFFORTS TO CURB USES.

87-922 E

CRS REPORT FOR CONGRESS

The tax-exempt bond provisions of the Tax Reform Act of 1986 represent a continuation of the congressional effort begun in the late 1960s to control the growth of State and local bonds issued for "private" purposes. This report describes the economic incentives that contributed to this growth and discusses the legislative efforts made to adjust these incentives from 1968 to 1986.

                                   by

The service makes such research available, without partisan bias, in many forms including studies, reports, compilations, digests, and background briefings. Upon request, CRS assists committees in analyzing legislative proposals and issues, and in assessing the possible effects of these proposals and their alternatives. The Service's senior specialists and subject analysts are also available for personal consultations in their respective fields of expertise.

                              CONTENTS

The tax-exempt bond provisions of the Tax Reform Act of 1986 represent a continuation of the congressional effort begun in the late 1960s to control the use of State and local bonds for "private" purposes. Several factors encouraged the growth of tax-exempt private-purpose bonds (also referred to as private-activity or nongovernmental bonds) during this period: (1) the absence of any practical guide for the Congress to use in the determination of what activities constituted a public purpose; (2) the understatement of costs to the State and local taxpayer caused by the absence of meaningful State and local sector financial responsibility; (3) the overstatement of the benefits from these bonds caused by the gap between benefits as seen by the State or local official and by the Federal policymaker; and (4) the private sector's awareness that use of tax-exempt bonds was a vehicle to reduce its borrowing costs.

This report explains how these factors encouraged the growth of private-purpose tax-exempt bonds. It then examines the efforts made by Congress to adjust these incentives by enacting dollar limits on bond volume to circumvent the difficulty of defining public purpose; raising costs to the State and local taxpayer; targeting recipients more carefully so that benefits are generated for the Nation as well as the State and local sector; and reducing the potential savings to private firms on their borrowing costs. The discussion of the legislation is organized by these four incentives rather than chronologically.

The report reaches several conclusions. Complete State and local freedom in bond issuance has proven to generate too many bonds at too high a Federal revenue cost. Controlling bond volume by imposing full financial responsibility for the bonds on State and local taxpayers might deny funding for some traditional public-purpose projects. And controlling volume by reaching a legislative consensus on what "private" activities also provide sufficient public benefits to merit tax exemption has proven to be elusive. The Tax Reform Act of 1986 pursued a middle path -- allowing State and local governments considerable freedom to choose their own projects while imposing a ceiling on each State's bond volume at a level consistent with Federal judgments on benefits from a national perspective.

INTRODUCTION

State and local governments, when raising capital, have usually favored the issuance of debt instruments whose interest income is exempt from Federal income taxation. Most of these debt instruments are in the form of long-term bonds. The volume of these long-term tax-exempt bonds grew rapidly during the last two decades, rising from an annual volume of $11.1 billion in 1965 to $204.0 billion in 1985. ¹ This rapid growth in bond volume was accompanied by a rapid increase in the Federal income tax revenue foregone from the interest income on these bonds. By 1986 the revenue loss on the outstanding stock of tax-exempt bonds amounted to an estimated $21.0 billion. From a different perspective, by the time the private-purpose bonds issued in 1985 are paid off, it has been estimated that the Federal revenue loss from their exempt interest income will be $58 billion. ²

The effort to control this revenue loss evolved into a struggle between two competing groups of public officials. One group, composed primarily of Treasury officials and some Congressmen, wanted to restrict the use of tax-exempt bonds to the financing of traditional State and local capital projects such as highways, schools, and public hospitals. The second group was composed primarily of State and local officials and some Congressmen. This group wanted to continue the usual practice of using these bonds for the financing of activities provided by private entities in cooperation with the State and local sector, such as industrial park development, student loans, and housing.

The result was almost continual congressional debate over tax exemption for bonds that have been called at various times private- purpose, private-activity, and nongovernmental. Numerous pieces of legislation were enacted, beginning with the Revenue and Expenditure Control Act of 1968 (Public Law 90-364) and continuing through the Tax Reform Act of 1986 (Public Law 99-514).

The first section of this report summarizes the growth of private-purpose bonds from the 1960s to today. The second section describes how four factors -- an inability to define public purpose, understatement of State and local taxpayer cost, overstatement of taxpayer benefit, and subsidy of private capital cost -- encouraged this growth. The third section describes the legislation that attempted to control this growth by adjusting or circumventing these incentives. The final section summarizes the evolution of the legislation and draws some observations from past experience for future legislative efforts.

THE GROWTH OF PRIVATE-PURPOSE TAX-EXEMPT BONDS

The growth in private-purpose tax-exempt bond volume since the 1960s is difficult to measure. Aggregation and reporting of data by activities consistent with the tax law's definition of public versus private purpose did not commence until reporting requirements were imposed by the Tax Equity and Fiscal Responsibility Act of 1982 (Public Law 97-248). However, some insight can be obtained by separating bond volume into general obligation (GO) bonds and revenue bonds. General obligation bonds pledge the taxing power of the political jurisdiction for payment of the bond proceeds and interest (debt service). Almost all GO bonds finance traditional State and local public infrastructure. Revenue bonds, which pledge for debt service the revenues of the project being constructed with the bond proceeds, are used to finance almost all private-purpose bonds as well as some indeterminate amount of bonds for traditional State and local infrastructure. Thus, revenue bonds can serve as a proxy for measuring trends in private-purpose bonds.

Figure 1 compares the volume of GO and revenue bonds and displays the revenue bond share of long-term bond volume from 1970 to 1984. The volume of both GOs and revenue bonds increased over the period. GOs increased from $11.85 to $24.53 billion, while revenue bonds increased from $6.1 to $65.87 billion. The revenue bond share of total long-term bond volume increased from 34 percent in 1970 to 73 percent in 1984. These data indicate there was substantial growth in the importance of revenue bonds and, presumably, private-purpose bonds.

"Presumably" is an important distinction, because these revenue bond totals also include bonds for nonprofit organizations and some traditional public purposes. But any lingering doubt about the growth of private-purpose bonds is removed by inspection of the data accumulated since Congress required in 1982 that State and local bond data be reported to the Treasury Department by private-purpose category. Figure 2 shows the volume and growth of new-issue bonds from 1983 to 1985 for the four major private-purpose categories: student loans, private exempt entities (nonprofit organizations), qualified mortgages, and industrial development. The volume of the most important category of private-purpose bonds, industrial development bonds (IDBs), rose from $27.1 billion in 1983 to $57.1 billion in 1985, an increase of 105.1 percent.

FIGURE 1. GENERAL OBLIGATION AND REVENUE BONDS: VOLUME AND REVENUE BOND SHARE

[Figure omitted]

FIGURE 2. PRIVATE-PURPOSE TAX-EXEMPT BONDS BY ACTIVITY: VOLUME AND GROWTH

[Figure omitted]

INCENTIVES AFFECTING THE GROWTH OF PRIVATE-PURPOSE BONDS

State and local officials responsible for bond issuance were beset by many different problems and constituencies that encouraged the issuance of tax-exempt bonds for private purposes. This section discusses four of the incentives: the difficulty in defining a public purpose; the understatement of costs to the State and local taxpayer; the overstatement of benefits to the State and local taxpayer; and the private sector's awareness that use of these bonds could reduce their cost of capital.

DEFINING PUBLIC PURPOSE

A major reason it was so difficult for Congress to control the use of tax-exempt bonds for private purposes was the difficulty of defining what constitutes a public purpose. In practice, determination of what activities satisfied a public purpose proceeded in two steps. State and local officials made the INITIAL decision to finance a particular activity with tax-exempt bonds, and then issued bonds for that purpose. This decision established the activity as a public purpose or, at least, as an activity eligible for tax-exempt financing.

Such State and local initiative was supported on both economic and legal grounds. First, it was consistent with the view that State and local officials are better able to discern their constituents' preferences for public goods than are more geographically remote Federal officials. Second, it was consistent with the view of many that the United States Constitution protects State and local governments against Federal intrusion in their financial affairs. In this view, the Constitution precludes the Federal Government from taxing the interest income from State and local bonds, whatever their purpose. ³

The second step in the definition of public purpose was the often-exercised congressional right to an after-the-fact review of these purposes. State and local officials tend to interpret these persistent efforts to limit the use of tax-exempt bonds for some activities as a congressional assertion that it believes itself to be a better judge of citizens' preferences for public goods than are State and local officials, and that Congress rejects the constitutional justification for tax exemption.

These differing views of congressional actions reflected an important fact -- although economic theory provides justification for Federal subsidy of the State and local sector, it is not a practical tool for the Congress to use to determine precisely what services should be subsidized. This point requires some explanation.

The economic justification for public goods provision suggests the State and local sector should provide goods and services when external benefits or costs preclude the private sector from providing the socially desirable amount. ⁴ This amounts to saying that goods provided publicly should have a substantial element of collective (or joint) consumption. It follows, then, that if the Federal Government is going to subsidize State and local provision, it should subsidize only those State and local goods with a strong element of collective consumption. Considerations of intergovernmental efficiency, however, suggest that not all collectively consumed goods provided by the State and local sector merit Federal subsidy. This more restrictive view would confine the Federal subsidy to those collective consumption goods that are likely to be underprovided by State and local governments.

Such underprovision results from the spillover of benefits among jurisdictions. The sheer number of State and local political jurisdictions implies that any one jurisdiction is likely to have a geographic reach that fails to encompass all individuals and businesses who benefit from its public goods. Thus, some of the collective consumption benefits spill over the border of a taxing jurisdiction, such as in the case of redistributive welfare programs, some educational services, or environmental projects. Collective consumption benefits from providing such goods exceed the benefits to taxpayers in the providing jurisdiction. Because many taxpayers are unlikely to be willing to pay for services received by others, it may be desirable for a higher level of government (which does receive tax payments from the spillover beneficiaries) to subsidize their consumption in order to induce State and local governments to provide the proper -- larger, amount.

The task of identifying those goods which possess externalities sufficiently numerous to merit State and local provision, but not provided in the proper quantities by the State and local sector, is akin to untying the Gordian knot. One's perception of externalities and the State and local versus Federal roles in their accommodation is undoubtedly influenced by all those noneconomic and nonquantifiable factors which determine our preferences, such as politics, religion, culture, and ethics.

A few examples are instructive in this regard, particularly because (as will be shown in the legislation section) these examples have been the subject of congressional action attempting to define what satisfies a public purpose. Golf courses, tennis courts, and other recreational facilities would be considered by many to be essentially private goods. But some individuals might perceive it to be unjust to withhold these goods from those incapable of paying for them. In effect, these individuals' sense of interdependence perceives benefits for all when those unable to pay gain access to the facilities. Should such individuals dominate the political process in a jurisdiction, golf courses and tennis courts become legitimate prospects for State and local provision. However, unless such preferences are representative of a fairly broad spectrum of voters in the United States, golf courses and tennis courts are not legitimate prospects for Federal subsidy of State and local provision. In this instance, the case for Federal subsidy fails due to the absence of spillovers.

Conversely, pollution abatement is often cited as an activity whose collective consumption benefits require public provision. It may be, furthermore, that the external costs from pollution caused by manufacturing activity spill out of the community in which they are produced, being carried by rivers to downstream communities or by air currents to adjacent communities. The citizens in the producing community are unlikely to provide sufficient incentives for pollution abatement to satisfy the citizens of these downstream or adjacent communities. If this pattern of external cost spillovers is repeated in a sufficient number of locations throughout the country, some Federal subsidy of State and local pollution abatement is likely to receive serious attention as is currently the case with acid rain and hazardous wastes. ⁵

The point here is simply to illustrate that there was no single answer to what goods and services the State and local sector should provide, and which of these should receive a Federal subsidy. Economic theory was quite useful in explaining the conditions under which Federal subsidy is desirable. But it was not easy to quantify these conditions in a manner which the Congress could use to determine public purpose, as will be made abundantly clear by the discussion in the next section of congressional attempts to define public purpose. The definition of public purpose is necessarily elastic and subject to continuous reexamination. In the political process this proved to be conducive to the growth of private-purpose bond volume.

UNDERSTATEMENT OF COSTS TO THE STATE AND LOCAL SECTOR

The extent of State and local taxpayers' financial responsibility for tax-exempt revenue bonds had a direct effect on State and local officials' perception of the cost to their constituents. State and local officials tended to view tax-exempt revenue bonds as being costless to their taxpayers, and thus to themselves. To understand this thinking it is necessary to be clear as to the distinction between a general obligation bond and a revenue bond. The GO bond pledges the tax revenues of the State or local government as payment for the debt service. A revenue bond usually pledges only the revenue stream generated by the project being built or revenue from other projects, but not the government's taxing power. ⁶ Almost all of the activities which have been the subject of congressional limitation efforts were financed with revenue bonds.

Beginning in 1968, the Internal Revenue Code did impose a relatively weak financial responsibility test for tax-exempt status of private-purpose bonds. If more than 25 percent of the bond proceeds was used in a trade or business, then not more than 25 percent of the debt service could be secured by the property or revenues of an entity engaged in a trade or business. This seemed to suggest bonds were eligible for tax exemption if the issuing entity (frequently an agency or authority created by a government) assumed responsibility for 75 percent or more of the debt service. But this simply meant the government pledged user charges or revenues from the project being built or from other projects. It did not necessarily mean that the State or local government must pledge to repay the debt service with tax revenues if these pledged project revenues were inadequate to cover debt service.

Should, for example, a hospital built with revenue bond proceeds be underutilized and revenues be inadequate for full payment of debt service, the issuing government was under no obligation to use tax revenues for the shortfall. Thus, the cost of an unsuccessful project fell upon bondholders in the form of defaulted or delayed payments. Because the other major cost of the subsidy, the foregone tax revenues from nontaxable interest income, fell upon Federal taxpayers, State and local government officials tended to act as though their middleman role generated no direct costs. ⁷

It is likely, however, that these tax-exempt revenue bonds imposed some indirect costs in the form of higher interest rates for general obligation financings. In theory, these private-purpose issues increased the supply of tax-exempt bonds and, assuming the supply of savings seeking tax-exempt shelter did not increase, should have raised the interest cost for financing traditional public sector infrastructure. In fact, there is some evidence to indicate that an increased supply of bonds did narrow the spread between tax-exempt and taxable bond yields. ⁸ In spite of this evidence, State and local officials tended to reject on minimize the importance of these overall higher interest costs.

A second factor in reducing the State and local taxpayer's cost was the investment of bond proceeds in higher-yielding taxable securities. These so-called arbitrage profits were frequently sufficient to pay not only for issuance costs but also to further reduce the effective interest rate on the bonds below the stated tax- exempt rate. ⁹

A simple example is helpful to understanding this technique. Assume a governmental unit wanted to construct a $10 million building. If it issued $10 million of bonds at a 5 percent interest rate, its annual interest cost would be $500,000. Absent any restrictions on arbitrage earnings, suppose the construction schedule enabled the governmental unit to invest a five-year average of $5 million of the proceeds in taxable securities yielding 7 percent. These interest earnings would amount to $350,000 each year. In effect, the governmental unit's actual interest expense is only $150,000 (its $500,000 interest expense minus is $350,000 interest earnings), an effective interest rate of 1.5 percent. Over the five year period the earnings amount to $1.75 million. This $1.75 million could be used to reduce the amount borrowed, to make the project bigger, or to reduce the governmental unit's tax rate and its taxpayers' tax bills.

A final aspect of the revenue bond bearing on its popularity and cost was the absence of a requirement for State and local officials to seek the explicit approval of their constituents for the issuance of these bonds. In many States, the constitutional and statutory restrictions imposed on bond issuance apply only to general obligation bonds, which pledge the taxing power of the jurisdiction to pay the debt service. A requirement to seek the approval of the general public implies that the public official must provide a justification for a bond issue, and is an obvious incentive to careful consideration of its costs and benefits.

In summary, the State and local sector did not see itself as bearing much of the cost of tax-exempt revenue bonds. Standard economic theory suggests that any good priced at zero (or even at less than its marginal cost of production) will generate excess demand. This suggests that the State and local sector issued a volume of bonds which on the margin cost society more than it valued the public (private?) goods being produced. Thus, another source of continual congressional attention to these bonds was the questionable level of inadequate State and local financial responsibility.

OVERSTATEMENT OF PUBLIC BENEFITS

State and local officials clearly saw private-purpose bonds as beneficial. First, they maintained that these bonds increased aggregate investment and Federal tax revenues, and were therefore beneficial national investments. Second, even if aggregate investment and Federal tax revenues did not increase, State and local government officials saw the bond proceeds as generating increased investment and jobs in the community, thereby expanding the State and local tax base. The question here is whether these officials' perception of "benefits" overstated the gains to society and created an incentive to issue too many bonds.

The first issue to address is the contention that these private- purpose bonds increased aggregate investment and employment. The ability of private-purpose revenue bonds to stimulate investment, employment, and tax revenues depends to a significant extent upon the responsiveness of savers to an increase in the interest rate. ¹⁰ Suppose savings are not responsive to a change in the interest rate. Then, private-purpose revenue bonds can be expected simply to reallocate investment among alternative uses. It may be true, as some contend, that these bonds generate Federal tax revenues and employment. But it is also true that Federal revenues and employment decline in those areas from which investment is displaced by the bonds. The net change is likely to be close to zero.

Alternatively, suppose savings are responsive to a change in the interest rate. In this case, private-purpose revenue bonds can stimulate investment and increase tax revenues, but so can numerous other investment subsidy programs. There is nothing unique about the stimulation properties of private-purpose revenue bonds. Utilizing a different investment stimulus program to achieve a macroeconomic goal makes no appreciable difference to the level of the Federal budget and the economy. ¹¹

Thus, from an aggregate perspective, whether or not private- purpose revenue bonds stimulated investment and generated Federal tax revenues is not important to an evaluation of the bonds. Any effect they had on investment and Federal revenues could have been provided by other means. In effect, the aggregate dollar value of outlays, revenues, and the deficit is a macroeconomic question. State and local officials' valuation of the national benefits of private- purpose revenue bonds was overstated to the extent this valuation assumed these bonds increase aggregate investment and Federal revenue.

The second issue to address is the contention that the choice of private-purpose revenue bonds as the instrument for investment stimulus made a difference to State and local governments. It is possible that those jurisdictions that used the bonds succeeded in generating more taxable economic activity than they would have received if, for example, liberalized depreciation rules were utilized to stimulate investment. But the evidence for this proposition is not very persuasive.

Ture, for example, purported to show that New York State's industrial development bond program produced many new jobs for the State. This may well have been the case, but the program depended upon the responsiveness of savers to a change in the interest rate (as Ture correctly noted). ¹² Given the mixed evidence concerning an increase in savings nationwide, it is possible that the jobs created or maintained in New York were offset by a reduction of jobs in other locations. In fact, Dewar suggested that Massachusetts's industrial development bond program did not produce many new jobs, as bond users reported they simply substituted the tax-exempt bond financing for other sources of capital. ¹³ In addition, Stutzer examined the proposition that small-issue industrial development bonds increased state wide employment or property tax base growth in Minnesota, and concluded that the bonds did not have a significant effect. ¹⁴

Thus, many evaluation studies suggest that even the State and local governments issuing the bonds did not receive appreciable benefits in some cases; and in cases where they did, it is likely that the increased investment and employment did not represent net benefits for the nation as a whole. What the search for "benefits" seems to come down to is that the DISTRIBUTION of investment and employment by location or socioeconomic group was affected. Unless this distributional effect was articulated by the Federal Government as a policy goal, however, one is hard pressed to see distributional changes as a "benefit" to society.

In addition, even if a distributional goal were articulated, it could not be achieved if the subsidy was available for firms generally regardless of location or socioeconomic group served. In such circumstances, no tax-induced capital cost differential exists and location choices are based on other factors. In fact, as will be seen in the next section, the congressional effort to restrict the volume of some private-purpose bonds by making eligibility for tax exemption dependent upon a more careful targeting of specific groups and locations does seem to suggest that Congress sees these distributional effects as a national benefit.

In conclusion, it appears that State and local officials probably overstated the national benefits of these bonds. As a consequence, State and local officials probably issued a volume of bonds that on the margin was likely to cost society more than it valued the public (private?) goods being produced.

REDUCTION OF THE PRIVATE COST OF CAPITAL

The only way a private firm or individual could obtain access to the low-cost financing provided by tax-exempt debt was through the cooperation of public officials. The greater was the potential reduction in the firm's cost of capital provided by tax-exempt financing, the greater was the pressure on State and local officials to issue more bonds on behalf of the nongovernmental user.

The Federal income tax law provided many incentives to stimulate investment. The incentive offered to the State and local sector was the low interest rate on its bonds (relative to the interest rate on taxable bonds) provided by the exclusion of interest income from Federal taxation. The incentives offered directly to the private sector were numerous, but the most important, prior to the Tax Reform Act of 1986, were the investment tax credit and accelerated depreciation (depreciation deductions in excess of economic depreciation). All of these incentives lowered the firm's rental cost of capital (the annual cost for a dollar of capital).

The use of tax-exempt financing for the capital needs of private firms or individuals enabled these private entities to combine the benefits of tax incentives designed for public and for private investment. Table 1 presents estimates of the rental cost of capital for a firm making an investment in 1982 in which the breakdown between equipment and structures is representative of the United States capital stock. The first column of row 1 shows the rental cost for an investment financed with 50 percent taxable debt and 50 percent equity which used the investment tax credit and accelerated depreciation allowances: the rental cost of capital was 13.13 percent.

The remaining columns of row 1 show what happens to the rental cost of capital when tax-exempt debt is substituted for taxable debt. The lower interest cost for tax-exempt debt caused the rental cost of capital to decrease to 12.95 percent when tax-exempt debt was used to finance 10 percent of the investment and to 12.07 percent when tax- exempt debt was used to finance 50 percent of the investment. Row 2 calculates the percentage reduction in the cost of capital from this use of the tax-exemption privilege. A project which substituted tax- exempt for all taxable debt (0.50 share financed with tax-exempt bonds) had its rental cost of capital reduced by slightly over 8 percent.

It is apparent from these estimates that the ability of a private entity to gain access to tax-exempt financing for the debt- financed portion of its investment significantly lowered its cost of capital. This provided a substantial incentive for the private sector to pressure public officials to act as conduits for such financing. Given the tendency of State and local officials to understate the costs and overstate the benefits to their taxpayers from issuing bonds on behalf of nongovernmental entities, this incentive probably helped to increase the supply of tax-exempt revenue bonds.

LEGISLATIVE EFFORTS TO RESTRICT THE GROWTH OF BONDS

The discussion in the preceding section provides a framework for understanding the major legislation enacted from 1968 to 1986 that attempted to control the use of tax-exempt bonds for private purposes. The discussion of this legislation is organized by the four types of economic incentives.

        TABLE 1. The Effect on the Rental Price of Capital of a

DEFINING PUBLIC PURPOSE

No effort is made here to present an exhaustive account of every legislative provision which can be interpreted as an effort to define what constitutes a public purpose. The major changes and some minor changes in the laws are described with an eye to giving the reader an appreciation for the complexity of congressional decisions over time as to what constitutes a public purpose.

The Revenue and Expenditure Control Act of 1968 (Public Law 90- 364) represented the first serious effort to attempt a definition of public purpose. In response to the growth of bonds issued for "private purposes," a two-part eligibility test was imposed that effectively denied tax exemption to most of the activities in question. Any bond that met a private use test AND a security interest test would be taxable. The private use test was satisfied if all or a major part (eventually defined as more than 25 percent) of the bond proceeds was to be used in a trade or business. The security interest test was satisfied if all or a major part (eventually defined as more than 25 percent) of the principal or interest was secured by or derived from property to be used in a trade or business. Note that these tests do not actually define a public purpose, but rather suggest that public purpose depends upon the type of entity that uses the proceeds and the security for the bonds. Bonds passing these two tests were classified as industrial development bonds (IDBs), and were not eligible for tax exemption. ¹⁵

Imposition of the use of proceeds test would have denied tax exemption to bonds issued for many activities the Congress deemed to serve a substantial public purpose. ¹⁶ For this reason, a lengthy list of exceptions to the tests was provided for specific activities, and it is the exceptions to these general eligibility restrictions which represent the first attempts to define public purpose.

Excepted activities were: (1) residential real property; (2) sports facilities; (3) facilities for a convention or trade show; (4) airports, docks, wharves, mass commuting facilities, parking facilities, or facilities for storage or training directly related to any of the foregoing; (5) sewage or solid waste disposal facilities, facilities for the local furnishing of electric energy, gas, or water; (6) air or water pollution control facilities; and (7) acquisition or development of land for industrial parks. In addition, any IDB issue of $1 million or less was excepted if the proceeds were used for the acquisition, construction, or improvement of land or depreciable property. At the election of the issuer, the $1 million limit could be increased to $5 million if the aggregate amount of related capital expenditures made over a six-year period was not expected to exceed $5 million.

Simply listing the exceptions does not do justice to the congressional effort at defining public purpose. One must read the committee reports and explanation of the conference agreement to appreciate the degree of legislative specificity being applied to the question of an activity's eligibility for tax exemption. ¹⁷ For example, the exception for residential real property was described as being for buildings containing one or more complete living facilities (whose components were defined) for non-transients, but which could include facilities for nonfamily purposes (such as a laundromat or other retail establishment); the sports facilities exception was to apply to both spectator and participation sports facilities, including such facilities as stadiums, ski slopes, golf courses, and facilities directly related to these exempt sports facilities.

The Revenue Act of 1971 (Public Law 92-178) made minor adjustments to public purpose definition. The amount of expenditures for unforeseen circumstances to be disregarded in the determination of the $1 million small-issue exception was increased from $250,000 to $1 million. The Act also eliminated the requirement that water facilities must be local (which relates to the geographic location of the users).

The Tax Reform Act of 1976 (Public Law 94-455) expanded the exceptions to the general eligibility restrictions. Tax exemption was extended to bonds issued by not-for-profit corporations organized by, or requested to act by, a State or local government for the purpose of acquiring student loan notes incurred under the Higher Education Act of 1965.

The Revenue Act of 1978 (Public Law 95-600) continued the liberalization of eligibility for tax exemption. The $5 million limit on six years capital expenditure for small-issue IDBs was increased to $10 million, and to $20 million for projects in certain economically distressed areas. Other adjustments were made to include a greater variety of projects in the exceptions for the local furnishing of electric energy and water facilities.

The Crude Oil Windfall Profit Tax Act of 1980 (Public Law 96- 223) further broadened the types of projects eligible for tax-exempt financing. Eligibility was extended to several types of solid waste disposal facilities, hydroelectric generating facilities, and renewable energy facilities. The Economic Recovery Tax Act of 1981 (Public Law 97-34) added to the list of exceptions by permitting exemptions for bonds used to finance mass commuting vehicles and the obligations of volunteer fire departments.

Beginning in the late 1970s, efforts were made to narrow the range of what is considered to be a public purpose. States and localities had begun to issue tax-exempt bonds to finance mortgages for single family owner-occupied housing. The Mortgage Subsidy Bond Tax Act of 1980, enacted as part of the Budget Reconciliation Act of 1980 (Public Law 94-449), placed income and purchase price restrictions on "qualified mortgage bonds," banned qualified mortgage bonds after 1983, and continued prior law for "qualified veterans' mortgage bonds."

The Tax Equity and Fiscal Responsibility Act of 1982 (Public Law 97-248) terminated the small-issue exception for bonds issued after 1986. In addition, it eliminated the use of small-issue IDBs beginning in 1983 if more than 25 percent of the proceeds were used for certain types of facilities -- automobile sales or service, retail food and beverage service (other than grocery stores), or provision of recreation or entertainment. No proceeds of exempt small-issues could be used for such activities as golf courses, massage parlors, hot tubs, and racetracks. Among numerous other actions, the Act relaxed some of the targeting restrictions on the issuance of mortgage subsidy bonds that had been implemented by the Mortgage Subsidy Bond Act of 1980.

The Deficit Reduction Act of 1984 (Public Law 98-369) continued the definition process. The sunset date for qualified mortgage bonds was extended to December 31, 1987. For the first time, restrictions also were placed on qualified veterans' mortgage bonds. Perhaps most important, an attempt was made to require the State and local sector to seek congressional review and approval in advance for all purposes for which tax exemption is available to nonexempt persons. Accordingly, tax exemption was eliminated if more than five percent of the proceeds of a bond issue were to be used for a purpose benefitting a nonexempt person and if that purpose had not been specifically approved by the Congress. The Act also extended the sunset date on small-issue IDBs for manufacturing facilities to December 31, 1986. Tax exemption was denied in some cases for bonds issued to finance the purchase of existing facilities and nonagricultural land. Finally, tax exemption was denied for bonds issued to finance any airplane, skybox or other private luxury box, health club facility, any facility primarily used for gambling, or any store the principal business of which is the sale of alcoholic beverages for consumption off premises. Additional restrictions were placed on student loan bonds.

Finally, the Tax Reform Act of 1986 further narrowed the definition of public purpose. The exceptions for sports facilities, convention and trade show facilities, parking facilities, and private pollution control facilities were eliminated. Other exempt facilities such as multi-family rental housing and airports were subjected to additional targeting and government ownership restrictions. A new exception was added for facilities to treat hazardous waste.

As this account illustrates, trying to define private activities that serve a public purpose is a tedious, complicated, and continual task in which the target moves over time. It is reasonable to view the congressional move to volume limitations on private-purpose bonds as a response to this frustrating process.

The first volume restriction was enacted in the Mortgage Subsidy Bond Act of 1980. The annual volume of qualified mortgage bonds in a State was limited to the greater of (1) nine percent of the average annual aggregate principal amount of mortgages executed during the three preceding years for single-family owner-occupied residences located in the state, or (2) $200 million. The Deficit Reduction Act of 1984 restricted qualified veterans' mortgage bonds to States that had programs in existence before June 22, 1984, and limited the volume to an amount equal to (1) the aggregate amount of such bonds issued by the State during the period from January 1, 1979 and June 22, 1984, divided by (2) the number (not to exceed five) of calendar years after 1979 and before 1985 during which the State actually issued qualified veterans bonds.

The Deficit Reduction Act of 1984 imposed a volume limitation on certain IDBs and student loan bonds set at the greater of $150 per State resident or $200 million. This represented an important step away from detailed congressional attention to public purpose definition. The object was to reduce the growth of private-purpose bonds while allowing the State and local sector to make decisions concerning what types of activities best serve public purposes and should be allocated part of the scarce private-purpose volume cap. Congress, however, retained some allocative control, and exempted some IDBs from the cap -- multifamily rental housing, convention or trade show facilities, airports, docks, wharves, and mass commuting facilities.

The Tax Reform Act of 1986 (Public Law 99-514) made the most important move away from public purpose definition. It set a volume cap for each State equal to the greater of $50 per capita or $150 million, effective in 1988. The only private-activity bonds not subject to the cap are those issued for nonprofit organizations; governmentally-owned airports, docks, wharves, and solid waste disposal facilities; and qualified veterans mortgages (which remain subject to their own cap). A cap of $150 million was also imposed on a nonprofit organization's outstanding stock of tax-exempt bonds, with an exception allowed for hospitals.

UNDERSTATEMENT OF COSTS TO THE STATE AND LOCAL SECTOR

Some might suggest that the Revenue and Expenditure Control Act of 1968 made the first attempt to correct the understatement of costs. The act imposed a financial responsibility requirement on State and local taxpayers for private-purpose bonds. If more than 25 percent of the bond proceeds was used in a trade or business, then not more than 25 percent of the debt service could be secured by the property or revenues of an entity engaged in a trade or business. This seemed to suggest bonds were eligible for tax exemption if the issuing entity (frequently an agency or authority created by a government) assumed responsibility for 75 percent or more of the debt service. But this simply meant the government pledged user charges or revenues from the project being built or from other projects. It did not necessarily mean that the State or local government pledged to repay the debt service with tax revenues if these pledged project revenues were inadequate to cover debt service. This change did not place taxpayers directly at risk to pay higher taxes, and as such probably did not raise taxpayers' cost evaluations appreciably.

The Tax Reform Act of 1986 strengthened this financial responsibility test. The security interest test is now met by 10 percent of the debt service. That is, tax exemption is conditioned on the State and local government being responsible for paying off at least 90 percent of the debt service. But it still does not require the tax base as security.

Legislation also attempted to increase costs to the State and local taxpayer by reducing the opportunities for earning arbitrage profits. Prior to 1969, the tax code had no restrictions on the ability of State and local governments to invest the proceeds of tax- exempt bonds in higher-yielding taxable securities. This placed the Federal Government in the position of providing an unintended source of revenue to State and local governments, since bonds issued strictly for arbitrage purposes generated no capital construction. The Tax Reform Act of 1969 (Public Law 91-172) provided for the taxation of State and local bonds where all or a major part of the proceeds can be reasonably expected to be used to acquire securities or obligations to earn arbitrage profits -- that is, to be used to acquire assets with a materially higher yield than that of the tax- exempt bonds. Allowances were made for State and local governments to earn some arbitrage profits on bond proceeds for temporary periods while awaiting expenditure for their primary purpose (capital construction), and for investment in a reasonable reserve or replacement fund not to exceed 15 percent of the total proceeds.

During the ensuing decade, numerous regulations were enacted to increase Federal control over the ability of State and local governments to earn arbitrage profits. The materially higher yield restriction was interpreted to mean a yield exceeding by more than one eighth of a percentage point the yield of the tax-exempt issue. In addition, State and local governments had taken to issuing term bonds rather than serial bonds. Sinking funds were created into which were placed tax or project revenues for repaying the principal of the term bonds at maturity. Since the regulations specified that it was the use of bond proceeds (not tax or project revenues) that was restricted for arbitrage purposes, the arbitrage regulations did not apply to these sinking funds. The monies in these funds were invested at unrestricted yield. This technique was curtailed by regulations issued in 1978 that required that all amounts held in a sinking fund, regardless of source, be treated as proceeds of the issue.

The Mortgage Subsidy Bond Tax Act of 1980 restricted to 1 percent the allowable spread between the interest rate on tax-exempt bonds and the interest rate on single-family home mortgages financed with the bond proceeds. It was estimated at the time that a 1.5 percent spread was necessary to cover a jurisdiction's administrative costs. A 1.5 percent spread was still allowed for multi-family housing. The Tax Equity and Fiscal Responsibility Act of 1982 increased the allowable spread for single-family housing to 1.125 percent.

The Deficit Reduction Act of 1984 required all arbitrage profits on IDBs to be rebated to the Treasury. The only exception was for issues in which all proceeds were expended within six months of the issue date and for the governmental purpose for which the bonds were issued. In addition, the Act restricted the amount of proceeds that could be invested at a yield above the tax-exempt bond yield to 150 percent of the debt service for the year, and required that the investment be reduced as the bond issue is repaid. Bonds issued for residential rental property were exempt from these restrictions, although still subjected to arbitrage restrictions that apply to all tax-exempt bonds.

The Tax Reform Act of 1986 extended the rebate requirement imposed on IDBs to all tax-exempt bonds. This requirement is waived if all proceeds of the bond issue are expended for the stated governmental purpose within 6 months. In addition, the restriction on the amount of proceeds that can be invested to 150 percent of annual debt service was extended to all private-activity bonds except exempt entities. These restrictions should reduce the ability of the State and local sector to lower its interest cost below the stated tax- exempt rate, raise the costs of bond issuance, and reduce the incentive to issue bonds.

TEFRA also addressed the issue of taxpayer approval for industrial development bonds. The issuing jurisdiction was presented with two options to qualify IDBs for tax exemption. First, a voter referendum could be held to approve the issue. Second, if the voter referendum alternative was not used, a public hearing must be held to inform the public of the proposed bond issue.

OVERSTATEMENT OF PUBLIC BENEFITS

The benefits from private-activity bonds were overstated because the investment and employment generated, although of value to the State and local governments, did not represent net increases for the Nation as a whole. Ultimately, it was decided that some private- activity bonds did generate national benefits provided that the investment and employment were targeted to specific locations or socioeconomic groups.

The evolving congressional interest in targeting benefits was most obvious in mortgage bonds. The Mortgage Subsidy Bond Tax Act of 1980 was designed to direct housing subsidies from tax-exempt bonds to those individuals who have the greatest need for the subsidy. For owner-occupied housing, mortgages were restricted to first-time home buyers (defined as not having an ownership interest in a principal residence for the three prior years); a price limit equal to 90 percent of the area median purchase price was imposed on bond- financed houses; and 20 percent of each bond issue was reserved for mortgages in targeted areas (defined either as a census tract in which 70 percent or more of the resident families have income that is 80 percent or less of the Statewide median family income, or as an area of chronic economic distress). For multi-family rental housing, at least 20 percent of the housing units was reserved for low or moderate-income tenants. This restriction was lowered to 15 percent for projects in targeted areas.

The Tax Equity and Fiscal Responsibility Act of 1982 relaxed some of the restrictions on single-family mortgages because of the distressed condition of the housing industry at that time. The three- year requirement for prior homeownership was required for only 90 (rather than 100) percent of the lendable proceeds. And the purchase price limitation was increased to 110 percent of median purchase price (120 percent in targeted areas). Minor adjustments were made to the targeting requirements for multi-family rental housing.

The Tax Reform Act of 1986 again changed the targeting requirements for single-family owner-occupied housing. The percent of the proceeds of a bond issue that must finance mortgages for home buyers satisfying the three-year prior ownership rule was increased to 95 percent. The purchase price limitation was decreased to 90 percent of median purchase price (110 percent in targeted areas). And an income limitation was imposed on mortgage recipients required family income not to exceed 115 percent of the higher of median family income for the area in which the residence is located, or the Statewide median family income. For multi-family rental housing, a requirement was instituted that 40 percent or more of the units in a project be occupied by tenants having incomes of 60 percent or less of the area median income, or that 20 percent or more of the units be occupied by tenants having incomes of 50 percent or less of the area median gross income.

Congressional concern with targeting can also be seen in the evolving treatment of bonds issued for redevelopment that are secured by tax revenues generated by the redevelopment. These were initially called "tax increment bonds." A jurisdiction would issue bonds to clear land or rehabilitate structures, rent or sell the redeveloped real property to the private sector, and use the incremental property tax revenues to pay debt service on the bonds. Since the bonds were secured by tax revenues, they qualified for tax exemption even though they had many of the characteristics of IDBs.

The Tax Reform Act of 1986 named such bonds "qualified redevelopment bonds" and restricted their use to specified "blighted" areas. A blighted area was defined as an area possessing excessive amounts of such things as vacant land on which structures were previously located, abandoned or vacant buildings, substandard structures, vacancies, and delinquincies in payment of property taxes.

REDUCTION OF THE PRIVATE COST OF CAPITAL

The ability of nongovernmental entities to combine the tax benefits designed for the private sector with the lower interest rate on tax-exempt bonds was curtailed somewhat by the TEFRA. The effect of this change is summarized in table 2.

The estimates in row 1 replicate those presented in row 1 of table 1, and represent the investor's rental cost of capital using tax-exempt financing with full depreciation allowances. TEFRA required the investor to choose to use either tax-exempt bonds with less generous accelerated depreciation allowances than permitted by current law or full depreciation allowances with no tax-exempt financing. The estimates in row 2 show the effect on the investor's rental cost of retaining tax-exempt financing and taking less generous depreciation allowances (straight line). The cost increased slightly, but to a level less than the 13.13 percent to which it would have risen if tax-exempt financing were relinquished (and full advantage were made of existing depreciation allowances). In other words, the accelerated depreciation benefits were worth much less to the investor than the benefits provided by tax exempt debt financing.

       TABLE 2. The Effect on the Rental Price of Capital of the

Another effort was made during the deliberations leading to the Deficit Reduction Act of 1984 to further restrict depreciation allowances for private entities using tax-exempt financing. This restriction would have required investors using tax-exempt financing to use both straight-line recovery AND a longer number of years for their depreciation allowances. This change would have reduced the financial incentive much more than the TEFRA restriction, but it was not adopted.

CONCLUSIONS

This account indicates that Federal legislation has focused on the economic factors which contributed to growth in the volume of tax-exempt bonds. The focus has changed steadily from defining public purpose in terms of specific activities to one in which non- traditional activities as a group are subjected to a quantitative limit: contrast the list of excepted activities in the Revenue and Expenditure Control Act of 1968 with the imposition of a strict volume limit on most non-traditional activities in the Tax Reform Act of 1986.

This evolution in approach seems to make sense. When one looks carefully at the list of private-purpose activities financed with tax-exempt bonds, such as economic development, student loans, owner- occupied and multi-family rental housing, exempt entity bonds (non- profit organizations), and even "small-issue" assistance to private businesses, it is striking that most of these activities also receive Federal support from the expenditure side of the budget or from other tax subsidies. ¹⁸ Congress has, by prior actions, implicitly certified some amount of most of these activities as serving public purposes. In fact, when one looks at the spectrum of activities in which the Federal Government is engaged, the Federal and State-local sectors' visions of public purpose seem markedly similar.

Attacking the growth problem with a fairly comprehensive volume limit accomplishes two goals simultaneously. First, it accommodates the diversity with which State and local governments divided responsibility between the public and private sectors. State and local officials can choose activities which best satisfy their vision of public purpose. Second, it enables the Federal Government to control its revenue loss.

Another evolution in legislative policy has been the increased use of targeting restrictions for activities such as housing. This reduces the dichotomy between Federal and State-local officials' perceptions of public benefits and reduces the likelihood that the State and local sector will overstate public benefits. The Congress has essentially said that benefit spillovers exist and Federal financial support is merited for some types of activities PROVIDED the benefits are targeted to people with certain income or demographic characteristics.

Legislation has also changed two of the financial incentives which prompt both State and local officials and private entities to issue too many bonds. First, arbitrage restrictions on bonds have decreased the ability of the State and local sector to reduce its interest cost below the stated tax-exempt interest rate. Second, the requirement to inform taxpayers of pending bond issues has increased the ability of taxpayers to evaluate their potential costs and benefits.

But the major understatement of costs for these bonds has not been addressed -- the absence of a requirement that the debt service on failed projects be paid by tax revenues. Abundant evidence exists that State and local taxpayers take very seriously the potential claims of outstanding debt obligations on their tax base -- most State and local governments have constitutional and/or statutory restrictions on the issuance of debt with claims against the tax base, usually in the form of debt ceilings and referenda requirements. A requirement that State and local governments carefully balance benefits and ALL costs of bond issuance by requiring tax base as the ultimate security for all bonds would reduce bond volume. Such a provision might even eliminate the need for volume caps and "use of proceeds" rules.

From the State and local perspective, their constitutional and statutory limitations on bond issuance would cause a problem in the short run. These limitations would have to be changed before more bonds backed by tax revenues could be issued. Resistance to changing these limitations is likely to be considerable. In fact, the limitations are at least in part responsible for the development of the revenue bond, which in many jurisdictions was a response to debt- limit levels which were inadequate to enable issuance of bonds even for some traditional infrastructure. Thus, caution is called for when considering changing the rules of the game at the Federal level.

The current compromise is probably a reasonable solution. Complete State and local freedom in bond issuance has proven to generate too many bonds at too high a Federal revenue cost. Imposing full financial responsibility on State and local governments might deny funding for traditional public purpose projects as State and local governments struggle to change their institutional arrangements. The middle path -- allowing State and local governments the freedom to choose their own projects while restricting the volume to a level consistent with Federal judgments on benefits from a national perspective -- has considerable appeal.

Should the allowable volume of tax-exempt private-purpose bonds prove to be inadequate and restrictions on incurring general obligation debt prove to be constraining, State and local officials have the option of issuing taxable bonds. This would require, of course, that State and local officials convince constituents that the benefits from the projects to be financed would be sufficiently great to compensate for the greater interest cost -- precisely the type of judgment Federal policymakers are trying to encourage.

² U.S. Department of the Treasury. Private Activity Tax-Exempt Bonds, 1985, by Phil Clark. Statistics of Income Bulletin. Vol. 6, Number 4. Spring 1987. pp. 43-50.

³ The legal justification for exemption rests on specific statutory exclusion of such income by Congress. Some claim that, absent such an exclusion, interest income would still be exempt as a matter of comity between the Federal and State and local governments, just as interest on Federal securities is exempt from State and local income taxation. The Supreme Court has never decided the question of the constitutional right of Congress to delete the statutory exclusion and tax the interest income. See U.S. Department of Treasury. Office of State and Local Finance. Federal-State-Local Fiscal Relations, Report to the President and the Congress, September 1985, p. 284.

⁴ This is the standard justification for government intervention in private markets which appears in every public finance text. For example, see Browning, Edward K., and Jacquelene M. Browning, Public Finance and the Price System, Macmillan Publishing Co., New York, 1983, (Chapter 2).

⁵ Debate continues on the desirable policy instrument to increase private pollution abatement. Regulation and/or taxation can also increase pollution abatement, but distribute the costs of abatement differently than a Federal subsidy of capital costs. The Tax Reform Act of 1986 has removed the tax-exempt status of bonds issued for private pollution control, but has inserted a new exception for hazardous waste treatment facilities.

⁶ One should not conclude from this discussion that revenue bonds were developed as a vehicle for diverting tax-exempt bond proceeds to the use of the private sector. The development of revenue bonds probably owes more to constitutional and legislative restrictions on state and local governments' ability to incur general obligation debt in relation to the size of their tax base. The revenue bond allows bonds to be issued which do not require referenda or do not count as part of a debt ceiling, because tax revenues are not pledged for payment. They have often been used for such public infrastructure projects as highways and bridges. See J.L. Bowers, Jr. Limitations on Municipal Indebtedness. Vanderbilt Law Review. 1951, pp. 37-52.

⁷ In discussing the reasons for the growth of mortgage revenue bonds for owner-occupied housing, the Congressional Budget Office states ". . . these programs appeal strongly to local officials who believe that the programs provide them with an opportunity to give positive benefits to their constituents with no expenditure of local or state tax revenues." Congressional Budget Office. Tax-Exempt Bonds for Single-Family Housing. Study prepared for the Subcommittee on the City of the Committee on Banking, Finance and Urban Affairs. House of Representatives. Committee Print 96-2, April 1979. p.5.

⁸ For a summary of the econometric estimates, see Peterson, G.E., J.A. Tuccillo, and J.C. Weicher. The Impact of Local Mortgage Revenue Bonds on Securities, Markets and Housing Policy Objectives. in G.C. Kaufman, ed. Efficiency in the Municipal Bond Market. Greenwich, Connecticut: JAI Press, Inc. 1981; and Peek, Joe, and James A. Wilcox. Tax Rates and Interest Rates on Tax-Exempt Securities. New England Economic Review. Federal Reserve Bank of Boston. Jan/Feb 1986, p. 29-41.

⁹ See Fabozzi, Frank J., et. al. Legal Arbitrage. In Frank J. Fabozzi, ed. The Municipal Bond Handbook. Homewood, Ill.: Dow Jones- Irwin, 1983.

¹⁰ The sensitivity of savings to the interest rate is a matter of considerable discussion. Most evidence seems to suggest savings are not very sensitive to interest rate changes, although some estimates do suggest a positive relationship that is significant. Some observers have suggested that, given the uncertainty surrounding this relationship, the most certain method of raising national savings is not through tax policy, but through raising government savings (reducing the deficit). See: vonFurstenburg, George M., and Burton C. Malkiel. The Government and Capital Formation, A Survey of Recent Issues. Journal of Economic Literature. September 1977; and E. Philip Howrey and Saul H. Hymans. The Measurement and Determination of Loanable-Funds Saving. In Joseph A. Pechman, ed. What Should Be Taxed: Income or Expenditure? Washington, D.C.: Brookings Institution, 1980.

¹¹ This Federal budgetary posture argument is not strictly true. Different sources of revenue and different spending programs will produce different incentives for business and individuals. Different incentives generate different responses in terms of altered economic decisions. The allocation of resources will differ, and output, employment, and revenues may differ. However, these "output" effects are sufficiently obscure that they are largely ignored in the budgetary process.

¹² Ture Norman B. Industrial Revenue Bonds: Estimates of Employment Effects and Size of Benefitting Companies. New York State Economic Development Council, Inc., 1983.

¹³ Dewar, Margaret E. The Usefulness of Industrial Revenue Bond Programs for State Economic Development: Some Evidence from Massachusetts. Joint Center for Urban Studies of the Massachusetts Institute of Technology and Harvard University. Working paper No. 63, March 1980.

¹⁴ Stutzer, Michael J. The Statewide Economic Impact of Small- Issue Industrial Revenue Bonds. Federal Reserve Bank of Minneapolis Quarterly Review, Spring 1985.

¹⁵ For a brief account of the historical development of IDBs, see Lamb, Robert and Stephen P. Rappaport. Municipal Bonds: The Comprehensive Review of Tax-Exempt Securities and Public Finance. McGraw-Hill, 1980. Chapter 11. Another review of State and local borrowing purposes from the 1740s to 1950 suggests that "the historical record is full of examples of state-local borrowing for what is now often called private-purpose tax-exempt borrowing." See Netzer, Dick. The Effect of Tax Simplification on State and Local Governments. In Federal Reserve Bank of Boston. Economic Consequences of Tax Simplification. Conference Series No. 29, 1985, pp. 222-251.

¹⁶ These exceptions were, in effect, are recognition that some flexibility in who uses the proceeds is consistent with economic theory. Direct government provision is only one alternative for altering the amount of an activity. The government could use the proceeds to build a facility and contract for private sector management. The government could also regulate or subsidize a privately owned and managed facility, or use the proceeds to subsidize private sector capital formation. The important point is that any of these alternative types of intervention can use tax- exempt bonds to reduce the cost of capital and increase output of a good or service judged to serve a public purpose. It does not matter whether the proceeds are used directly by the government, or by businesses, individuals, and non-profit entities. The choice among alternative agents (users of the proceeds) should depend upon their relative cost in altering output. In principle, flexibility in who uses bond proceeds is desirable.

¹⁷ U.S. Congress. Revenue and Expenditure Control Act of 1968: Explanation of the Bill H.R. 15414 As Agreed to in Conference. Conference Committee Print. 90th Cong. 2d sess. June 10, 1968. U.S. Govt. Print. Off.

¹⁸ A concise summary of the financial assistance provided to these activities can be gleaned from any edition of: Office of Management and Budget. Special Analyses: Budget of the United States Government. The chapters on Federal credit programs, tax expenditures, and aid to State and local governments are of particular interest.