Congressional Research Service • Library of Congress • Washington, D.C. 20540

Statement Submitted for the Hearing Record

Senate Finance Committee

March 3, 1999

Tax-exempt Bond Proposals to Increase

Public Elementary and Secondary School Facilities

Dr. Dennis Zimmerman

Specialist in Public Finance

Congressional Research Service

Federal Financial Support and Policy Objectives

State and local governments historically have assumed most of the financial responsibility for public elementary and secondary schools. They raised about 92 percent of total school revenue for school year 1995-96; the federal government contributed about eight percent of revenue.

Federal financial support can be divided into two major components. Direct federal support provided by on-budget spending programs in school year 1995-96 amounted to $19.1 billion (as measured by the states), 6.6 percent of total school revenue. The federal policy objectives of this direct federal spending are fairly clear: 55 percent of this assistance in fiscal year 1995 targeted disadvantaged children; another 22 percent targeted disabled children; 12 percent targeted school system support for such things as professional development and drug abuse education; and six percent targeted children whose parents live and/or work on federal property.⁽¹⁾

Indirect federal support for capital facilities is provided through the tax system. The interest income individuals and businesses earn on state and local debt is excluded from their taxable income. This exclusion lowers the interest rate on state-local debt, a reduction in effect paid for by the federal tax revenue not collected on the excluded interest earnings. The estimated revenue loss on school facilities bonds amounted to $3.7 billion in 1996, about 1.2 percent of total education revenue.⁽²⁾ The federal government imposes no limit on the amount of tax-exempt bonds state-local governments may issue for governmentally owned school facilities.

Unlike federal direct spending for public elementary and secondary schools, this tax subsidy is not motivated by a federal education policy objective. Its existence is a byproduct of the income tax structure established in 1913 which incorporated the concept that the various levels of government should refrain from taxing each other. As a result, the tax subsidy is identical for all state-local capital facilities -- schools, roads, hospitals, parks, etc. -- and does not affect state-local taxpayer choices among different types of facilities.

In summary, three facts stand out about federal financial support for public elementary and secondary schools:

It is minor compared to state-local support.

On-budget spending is targeted to four major policy objectives (the disadvantaged, the disabled, system support, and the federally impacted).

The major tax subsidy was not adopted to pursue a federal education policy objective, and has been structured not to influence state-local taxpayer choice among capital facilities for different public services.

The State-Local Sector and America's Public School Facilities

Attention recently has focused on the deficiencies of public elementary and secondary school capital facilities. Studies have suggested that as much as $112 billion of investment may be necessary to restore school facilities to good overall condition, and that the resources of many local school districts are inadequate to rectify the situation.⁽³⁾

It is useful to evaluate this information in an economic context. The gap between "good overall condition" of school facilities and their current condition is a serious problem not to be minimized that undoubtedly has an adverse impact on human capital formation. But budget constraints are a fact of life: our desire for both private and public spending (consumption) exceeds our ability to pay for it. It is likely that a similar study assessing the condition of state-local capital facilities for any function -- roads, sewage treatment plants, prisons -- would reach a similar conclusion.⁽⁴⁾ A gap exists between the "good overall condition" of the capital stock we desire and the less-than-good overall condition we choose to live with.

When making budget allocation decisions, state-local decision makers decide where to spend additional tax revenue based in part upon their assessment of which activity will provide the highest return or value. It is a given that positive returns will result from additional investment in almost any activity funded by state-local budgets. But a ten percent return in education facilities will not be funded if decision makers judge a twelve percent return is available in sewage treatment facilities. In other words, one must consider the possibility that state-local decision makers made their spending decisions with complete information; that they chose the existing less-than-good condition of education facilities because they place a higher value on spending the available tax revenue for private consumption or other state-local services.

For the Nation as a whole, state-local taxpayers have not been neglecting education facilities. Table 1 presents referendum data on public elementary and secondary school bond issues for the years 1988 through 1998. The percentage of bond issues approved and the percentage of dollars approved appear in columns 2 and 3. Both series tell approximately the same story. Approval rates declined substantially in the early 1990s, reaching a low of 49.9 percent for Issues in 1991 and 48.4 percent for Dollars in 1993. Since those lows, the approval percentage for both Issues and Dollars has risen substantially. The 1998 approval rates of 66.8 percent for Issues and 82.4 percent for Dollars are now higher than the levels that prevailed in 1988.

The increasing approval rates are consistent with the 7.7 percent real annual growth rate of school bond volume (dollars of new issues) that occurred from 1979 through 1998. This is not surprising. We are now in the longest uninterrupted economic expansion in the Nation's history, during which the state-local surplus rose from $80.1 billion in 1990 to $148.7 billion in 1998. As real income rises, state-local taxpayers can be expected to spend more on a wide range of public services, including investment in schools. But these bond data do not provide evidence about how much of the growing bond volume was necessary to keep pace with growing student enrollment and whether schools were faring better or worse than other state-local services.

Table 2 compares the 7.7 percent real annual growth rate of school bond volume over the last two decades to the rates for school-age population (ages 5 to 19) and state-local receipts net of federal grants.

The school-age population grew at a 0.2% annual rate, so most of this 7.7 percent real annual increase in bond volume was devoted to maintaining or improving the facilities of a relatively stable school population. State-local receipts net of federal grants grew at a 4.1 percent real annual rate. These data suggest state-local taxpayers have been devoting an increasing share of own-financed revenue to schools, and school construction spending has fared better than all other functions combined.

Of course, these aggregate data undoubtedly mask a considerable amount of variation among states and school districts. Several circumstances arise which may cause school districts to provide grossly inadequate school facilities, and alleviation of some of these circumstances may be consistent with historical federal policy objectives for financing public elementary and secondary education.

A district might suffer from inadequate fiscal capacity; residents may be poor and the district may lack significant commercial and industrial property tax base. If its state does not have a vigorous fiscal equalization program for education finance, resources may not be available to provide minimal capital facilities.

Some school districts might experience a substantial influx of retirees, or be at the height of a long-term aging of their population. Retirees may feel they have done their duty by supporting school finance in their child-raising years. Seeing few direct benefits to themselves, they may be reluctant to support additional spending to maintain minimal services, particularly if they have relocated.

Some school districts have experienced rapid population growth (often resulting from immigration to the United States). A "normal" financing effort might prove to be inadequate to maintain minimal services when student enrollment expands rapidly.

Some states and local governments impose very tight borrowing restrictions and/or super-majority approval requirements for bond referenda that may frustrate the majority's spending preferences.

In summary:

The condition of America's school facilities may or may not be worse than the capital facilities for other state-local public services.

The proportion of school bond votes approved rose from a low of 50 percent in 1991 to 67 percent in 1998. The percentage of dollars approved in 1998 was 82 percent versus 49 percent in 1991.

State-local taxpayers have devoted an increasing share of their own-source revenue to school bond finance; over the last twenty years, the volume of new-issue school bonds has grown at a 7.7 percent real annual rate, while state-local own-source revenue has grown at a 4.1 percent real annual rate. Since the school-age population has grown at a mere 0.2 percent rate, most of this spending has been devoted to maintaining or improving facilities.

These data present a favorable picture for the Nation's school facilities, but may hide a subset of communities that find it difficult to maintain adequate school facilities due to: a high concentration of the poor; a concentration of retirees who are reluctant to support school spending; high population growth rates, sometimes resulting from an influx of immigrants; and very tight borrowing restrictions and/or super-majority requirements for approval of bond referenda.

Tax-exempt Bond Proposals

Several proposals have been introduced that would adjust the current tax treatment of state-local debt to increase federal financial support for school construction.⁽⁵⁾ The Administration has proposed Tax Credits for Holders of Qualified School Modernization Bonds and Qualified Zone Academy Bonds; Representative Archer has proposed a lengthening of the period during which arbitrage can be earned and not rebated to the Treasury; Senator Graham has proposed allowing school facilities to be financed with private-activity bonds; and it has been proposed that the annual issuance ceiling to qualify for the small-issuer arbitrage rebate exemption be raised. The last two proposals were adopted by the Senate Finance Committee but not accepted by the Conference.

Each of these proposals is described. Each proposal's effect on the share of the debt service costs borne by state-local taxpayers is estimated, and the targeting of the proposal is compared to the targeting of federal on-budget spending for elementary and secondary education.

School Modernization Bonds.

Description. This Administration proposal would authorize issuance of $11 billion of tax credit bonds in 2000 and $11 billion in 2001. School bond volume in 1998 was about $23 billion, so this proposal could be available to approximately 50 percent of the school bond market in 2000 and 2001.

Cost Reduction. Tax credit bonds pay 100 percent of state-local interest cost on bonds, as opposed to 25 to 30 percent of interest costs for traditional tax-exempt bonds. Thus, unlike tax-exempt bonds, tax credit bonds lower the cost of investing in school facilities relative to investing in capital facilities for any other public purpose. This lower relative cost would be a powerful incentive for state-local taxpayers to adjust their public budgets and provide more education services and less of all other services.

Targeting. Half of the annual borrowing authority would be reserved for the Nation's communities with the highest incidence of children living in poverty. The remaining half would be allocated to the states and qualifying school districts based upon the federal assistance they received under the Basic Grant Formula for Title I of the Elementary and Secondary Education Act of 1965 (based primarily upon incidence of low-income children). But states would not be constrained by the Title I formula and could use any appropriate mechanism for distributing the funds. Thus, half of the subsidy would conform to the federal government's existing criteria for federal spending programs in education, and half could potentially be spent on other school districts.

Relaxation of Arbitrage Restrictions.

Description. State-local arbitrage bonds are tax-exempt bonds issued where all or a major portion of the proceeds are used to acquire securities with a higher yield. Because state-local governments pay no federal income tax on their interest earnings, Congress has restricted their ability to earn arbitrage profits. Bonds for construction are allowed to earn arbitrage profits if they conform to a schedule for spending the bond proceeds: 10% within six months of issuance; 45% within 12 months of issuance; 75% within 18 months of issuance; 95% within 24 months of issuance; and the permissible 5% retainage (amounts by which the earlier targets are missed) within 36 months. Failure to comply triggers a requirement to rebate the arbitrage earnings to the U.S. Treasury.

This proposal would slow and lengthen the spend-down schedule that must be met for bonds issued to finance public school education facilities in order to qualify for exemption from arbitrage rebate. No rebate would be required if: 10 percent of bond proceeds is spent within 1 year of issuance; 30 percent is spent within 2 years; 50 percent is spent within 3 years; and 95 percent is spent within 4 years. The 5 percent retainage would have to be spent within 5 years. The proposal applies to all school bonds.

Cost reduction. Issuers must be cautious when attempting to earn arbitrage profits. Suppose the interest rate on the tax-exempt bond issue is 6 percent and the interest rate on a comparable long-term taxable bond is 8 percent. In theory, the issuer could earn 2 percent arbitrage profit by investing the tax-exempt bond proceeds in 8 percent long-term taxable securities. This is a risky investment strategy. The issuer's investment horizon is short because the spend-down rules require sale of all the securities within 36 months (60 months if this proposal is passed). Should interest rates have risen when the issuer must sell the taxable bond to pay for construction costs, the bond must be sold at a discount and the issuer will suffer a capital loss that could easily exceed the arbitrage earnings. Thus, the calculations in this testimony assume the issuer earns arbitrage profits of 0.75 percent, not the 2 percent yield differential. The important point here is not so much the share of the principal that could be paid off by the arbitrage profits, but the differential between current law and the proposed changes.

Assuming the issuer takes maximum advantage of arbitrage opportunities with a 0.75 percent profit, current law could provide arbitrage profits for tax-exempt bonds sufficient to pay for 1.05 percent of the amount borrowed. For tax credit bonds, this percentage would rise to 9.5.⁽⁶⁾ Allowing a five-year spend-down period for tax-exempt bonds would increase the percentage borrowed that could be financed with arbitrage profits from 1.05 to 2.4 percent. If combined with tax credit bonds, the percentage would rise from 9.5 to 21.2 percent.

Targeting. The arbitrage proposal would apply to all school bonds. No attempt is made to target its availability to school districts that meet the federal government's targeting criteria for its on-budget spending programs.

Public School Construction Partnership Act.

Description. This proposal introduced by Senator Graham in the 105^th Congress would include public elementary and secondary education facilities in the list of exempt facilities eligible for the use of tax-exempt private-activity bonds. A state could issue bonds equal to the greater of $10 per resident or $5 million on behalf of corporations that would use the bond proceeds to build school facilities and lease the buildings to school districts. A corporation must charge a lease payment such that the building could be transferred to the school district at the end of the contract without further compensation to the corporation. The bonds would not be subject to the private-activity bond volume cap, so they would not compete with other private-activity bonds for scarce borrowing authority.

Cost reduction. This proposal might reduce the federal subsidy. Private-activity education facility bonds would be issued as revenue bonds whose debt service is secured by the corporation building and operating the facility rather than as general obligation bonds whose debt service is secured by the full faith and credit of the issuing school district. As a result, the interest rate on the private-activity school bonds is likely to be higher and the spread between the taxable interest rate and the interest rate on the school bonds is likely to be lower. The federal government would pay a smaller share of interest costs than it would pay on governmental tax-exempt school bonds.

A school district that chose this option could conceivably receive compensation sufficient to offset its higher interest cost in two ways. First, it might face very restrictive bond referenda requirements that preclude getting approval from the voters. Although private-activity bonds require the issuing jurisdiction to hold a public meeting, they do not require a vote. Second, the corporation might be a more efficient builder and operator of the facility, or it may be able to avoid compliance with a host of regulatory rules pertaining to government construction projects (such as the Davis-Bacon Act). These savings might enable the corporation to provide lease terms whose present discounted value is lower than would be the case for principal and interest payments on the debt.⁽⁷⁾

Targeting. All but $5 million must be allocated to high-growth school districts, defined as having: (1) a 5,000 or greater student enrollment in the second academic year preceding the date of the bond issuance; and (2) an increase in student enrollment of at least 20 percent in the 5-year period ending with that second academic year. It is not clear how many of the eligible districts would have characteristics that are targeted by federal on-budget education spending.

Small Issuer Arbitrage Exemption.

Description. When the requirement for rebate of arbitrage earnings was enacted in 1986, governmental units that issued no more than $5 million of bonds per year were exempt. In 1997, the exemption limit was increased to $10 million, provided at least $5 million is used to finance public school construction. This proposal would increase the exemption limit to $15 million, provided at least $10 million is used to finance public school construction.

Cost reduction. The value of the small-issuer exemption is that the spend-down rules do not apply; the issuer can earn arbitrage profits on the amount borrowed for the entire three-year spend-down period. When considering a $5 million marginal investment on a variety of public functions, state-local taxpayers will likely notice that (under current law) school bonds could earn arbitrage profits sufficient to pay 2.3 percent of the amount borrowed, while bonds for other functions could earn arbitrage profits sufficient to pay only 1.05 percent of the amount borrowed. If tax credit bonds could be combined with the small-issuer exception (while retaining the three-year spend-down requirement), arbitrage profits would be sufficient to pay 20.3 percent of the amount borrowed.

Targeting. This provision would apply only to relatively small governmental units. It is not clear how many of these units would have the characteristics that are targeted by federal on-budget education spending.

1. U.S. Library of Congress, Congressional Research Service, Public School Expenditure Disparities: Size, Sources, and Debates over Their Significance, No. 96-51 EPW by Wayne Riddle and Liane White, December 19, 1995, 31p.

2. Indirect financial support is also provided by the deductibility of state-local income and property taxes from federal taxable income. This provision is not discussed here. The tax-exempt bond revenue estimate is based on a 1996 federal revenue loss from all outstanding bonds of $25 billion (Budget of the U.S. Government, Analytical Perspectives, Fiscal Year 1998), and assumes the school share of the outstanding stock of all state-local bonds is equal to the school share (14.7 percent) of new-issue state-local bonds issued in 1996. A small amount of tax credit bonds are also available for school districts with high concentrations of students receiving free lunch.

3. U.S. General Accounting Office, School Facilities: America's Schools Not Designed or Equipped for 21^st Century, GAO/HEHS-95-95, April 4, 1995; and GAO, School Facilities: Condition of America's Schools, GAO/HEHS-95-61, February 1, 1995.

4. For an example, see Commission to Promote Investment in America's Infrastructure, Financing the Future: Report of the Commission to Promote Investment in America's Infrastructure, February 1993.

5. The question of whether these proposed increased federal subsidies represent an improvement in economic efficiency is complex. The answer depends in part upon the extent to which returns from elementary and secondary education accrue to society rather than the individual and how widely these "external" benefits spill beyond state borders.

6. Since the federal government pays 100 percent of the interest cost on tax credit bonds, arbitrage earnings would be 6.75 percent, not the 0.75 percent for tax-exempt bonds.

7. Some have suggested the efficiencies in such public/private partnerships may be sufficiently great that school districts could reduce costs even if they used taxable debt. Ronald D. Utt, How Public-Private Partnerships Can Facilitate Public School Construction, Heritage Foundation Backgrounder No. 1257, February 25, 1999.