Replace the Tax Exclusion for Interest Income on State and Local Bonds with a Direct Subsidy for the Issuer

The federal tax code offers several types of tax preferences for bonds issued by state and local governments. First, interest income from governmental bonds--those that state and local governments issue to finance their own activities--can be excluded from a bondholder's adjusted gross income (AGI). As a result, the interest earned on those bonds is exempt from the federal income tax. Second, the interest earned on qualified private activity bonds--which states and localities can issue to finance certain types of socially beneficial projects undertaken by private entities, such as the construction of hospitals and schools--can also be excluded from a bondholder's AGI. (In some cases, however, that income may be subject to the alternative minimum tax.) As a result of those tax preferences, borrowers pay less interest than they would on comparable bonds with taxable interest. The revenue forgone by the federal government effectively pays part of the borrowing costs of state and local governments and of those private entities whose bonds qualify for a tax preference.

For governmental and qualified private activity bonds issued in 2012 and afterward, this option would replace the exclusion for interest income with a subsidy paid directly to the issuers of the bonds. Under the option, borrowers would make taxable interest payments to bondholders; in turn, state and local issuers of those bonds would receive a subsidy payment directly from the federal government equal to 15 percent of the interest paid on the bonds. (States and localities that issued qualified private activity bonds would pass the subsidy payment on to the corresponding borrower in the private sector.) The option would retain restrictions that apply to governmental bonds, such as those on arbitrage earnings. In addition, it would not alter any of the limits--such as volume caps--currently imposed on the issuance of qualified private activity bonds. If implemented, such a policy change would increase federal revenues by more than it would increase federal outlays, for a net saving of $31 billion from 2012 through 2016 and of $143 billion from 2012 through 2021.

Making subsidy payments to borrowers for the bond interest they pay could have several advantages. First, switching from a tax exclusion to a subsidy payment for bond interest would be a more cost-effective way of providing a subsidy to borrowers. Only a portion of the federal revenue currently forgone through a tax exclusion on bond interest income actually lowers financing costs; in contrast, borrowers would receive all of the benefits of direct subsidy payments for bond interest. The tax exclusion on interest income is less cost-effective because some of the federal revenue forgone through the tax exclusion goes to bondholders in higher tax brackets. They receive gains that exceed the investment return necessary to clear the market for such bonds--that is, to attract enough buyers so that the demand for bonds matches the amount supplied. (In order to attract sufficient buyers in lower tax brackets to sell all of the bonds, issuers have to offer higher yields than are necessary to attract at least some of the buyers in higher tax brackets.)

Second, the amount of the federal subsidy for borrowing would be explicit and unaffected by other federal policy decisions. Currently, for example, the savings in financing costs that are realized by issuing governmental or qualified private activity bonds are largely determined indirectly by other features of the federal tax code (such as bond buyers' tax brackets, which are an important determinant of the demand for such debt).

Third, making subsidy payments to the issuers of bonds could improve federal budgeting practices. The subsidies provided under current law are not readily visible in the budget; in contrast, under this option, the federal government would know the exact amount of financing subsidies it was providing in a given fiscal year. As a result, policymakers would be able to accurately assess the cost of the subsidies in comparison with the cost of other types of assistance to state and local governments and other spending more generally.

A disadvantage of the option is that it could raise borrowing costs for issuers of tax-preferred debt and thereby deter some investment that might have national benefits or place greater burdens on already strained state and local budgets. A subsidy payment rate of 15 percent is at the lower end of the range of estimated annual reductions in interest payments attributable to the exclusion of bondholders' income from federal taxes. Thus, state and local governments and some private entities would probably have to pay more to borrow, which in turn could cause them to reduce their spending on capital projects, such as schools and roads, and on existing programs (because general tax revenues are often used both to pay principal and interest on debt and to fund ongoing government operations). Borrowing costs could rise even further if lenders became fearful that transforming the subsidy from a tax exclusion to a more visible expenditure program would lead to further reductions (and potentially elimination) of the subsidy in the future and, as a consequence, charged higher rates of interest.

However, the option could increase or decrease demand for tax-preferred debt, and it is therefore unclear whether borrowing costs would fall or rise. On the one hand, offering a taxable--and thus higher--yield on bonds issued by state and local governments would make that debt attractive to new types of bond buyers (such as managers of pension funds and foreign investors, who generally do not pay U.S. taxes). The demand created by those purchasers could lower borrowing costs. On the other hand, the demand for tax-preferred bonds could fall as a result of converting the tax exclusion on interest income into a direct subsidy payment. That is because the option would reduce the after-tax returns on such bonds for people who are in higher tax brackets (although the yield would be higher, they would pay taxes on all of the interest income they received) and, as a result, could lead them to buy fewer of those bonds.

If a 15 percent credit were to result in higher borrowing costs that deterred investments considered desirable by the federal government, the amount of the subsidy payment could be increased, either overall--say, from 15 percent to 20 percent or 25 percent--or in a way that depended upon the purpose for which a bond was issued. However, increasing the subsidy rate would yield less savings for the federal government.