Unlock tax planning solutions to

maximize your investment results.
achieve your financial goals.
comply with regulations.

Leverage a proprietary dataset from 190+ countries, Artificial Intelligence, and a network of top local experts.

Repeal Tax Exemption for Municipal Bonds


The proposal would repeal for future years the exemption for interest on state and local bonds. While the exemption is supposed to help state and local borrowers, only a fraction of the federal cost of the exemption is captured by the borrowers. What is captured, in the form of lower interest paid, does harm by inducing public ownership of projects that should not be undertaken. For outstanding bonds, the interest would be taxable in future years, but holders would get a tax credit in the amount of the reduced interest they have accepted from municipal bonds. State and local governments would get a direct subsidy exceeding their benefit under current law.

Current Law

Under section 103, gross income does not include interest on state or local bonds. Most (83 percent) of the bonds that are tax exempt under section 103 are general obligation bonds in which the proceeds are used by the state and local governments and repayment is from general tax revenues.[1] However, 9.4 percent of tax- exempt section 103 bonds are issued on behalf of non- profit organizations for specified qualified purposes,[2] and 7.3 percent of section 103 bonds are private activity revenue bonds, issued for the use of private businesses.[3]

For private activity bonds, the state or local government is merely a conduit and repayment is available only out of the revenue provided by the true private borrower.

Eligibility for section 103 tax-exempt borrowing, be- yond general obligation bonds, is specified by statute with great complexity. Among the specified borrowings eligible under section 103 are some voluntary fire departments (section 150(e)), qualified scholarship funding bonds (section 150(d)), private activity bonds that are exempt facility bonds, qualified mortgage bonds, veterans’ mortgage bonds, small issue bonds, qualified student loan bonds, qualified redevelopment bonds, and qualified section 501(c)(3) bonds. Section 141(e). Private activity bonds must meet a volume cap on how much any one state or state agency can offer. Section 146. However, Congress has exempted some private activity bonds from the volume caps, including, for instance, borrowing for airports,[4] for docks and wharves,[5] and for construction in the Liberty Zone in New York City.[6] Interest from private activity revenue bonds is now subject to alternative minimum tax of up to 28 percent for individuals and 20 percent for corporations, unless the borrowing is by a charitable organization for its charitable purpose.[7]

A qualified borrowing cannot be an arbitrage bond, that is, a bond expected to be used to purchase an investment yielding a rate higher than the interest on the section 103 bonds, with specified exceptions.

A prospective issuer may ask for a letter ruling on qualification for exemption (Rev. Proc. 96-16, 1996-1 C.B. 630) and may appeal an adverse ruling (Rev. Proc. 99-35, 1999-2 C.B. 501).

Interest incurred to buy or carry tax-exempt bonds is matched to the exempt income and is, therefore, not deductible. Section 265(a)(2). Allocation of the interest expense to the exempt income is a complicated and ineffective way to prevent deduction of extra interest that would be avoided if the exempt bonds were not carried. The combination of exempt income and deducted costs, when allowed, allows unrelated income to be sheltered from tax.

Interest paid by the federal government on its own bonds is not exempt from federal income tax.

Reasons for Change

The purpose of section 103 is to subsidize some borrowing, but the exemption wastes most of the federal cost because the intended beneficiaries cannot capture the cost. The exemption for interest on state and local bonds arose, not from a well engineered subsidy program, but because of doubts about the constitutionality of a tax on state bonds.[8] It has now been settled that Congress may tax, on a nondiscriminatory basis, interest from whatever source derived.[9] With the disappearance of the constitutional grounds, the only remaining ground for the exemption is its delivery of a subsidy to qualified borrowers. On that ground, justification for the exemption fails because state and local borrowers capture so little of the federal cost and because the borrowing should not be subsidized without budget-imposed discipline.

Delivery efficiency of tax exemption. If a municipality could sell its bonds only to investors in the 35 percent bracket and they had no tax-advantaged alternatives, the municipality could capture whatever the federal government lost. If prevailing taxable interest were 10 percent for some term and risk rating, for instance, municipalities offering tax-exempt bonds would need to pay only a jot over 6.5 percent interest to attract lenders because taxable investors in a 35 percent bracket would get only 6.5 percent from their taxable bonds alternative after tax. Under those circumstances, the federal government would lose 35 percent of prevailing taxable interest by forgoing tax, and the municipalities would achieve a benefit of almost 35 percent, by a discount or drop measured from the same prevailing interest. The federal cost and the state and local benefit, both measured from the baseline of interest on taxable bonds, would be nearly the same.

The difficulty with the exemption is that there are too many tax-advantaged alternatives offered on the market so that state and local borrowers cannot attract any investors who will accept a 35 percent discount. Currently, the discount on long-term AAA state and local bonds is between 2 percent and 9 percent.[10] Long-term bonds have the lowest discount because taxable investors find it easier to avoid tax on competing long-term investments than on shorter-term investments. The sup- ply of tax-advantaged investments floods the market. State and local long-term borrowers must offer not 65 percent of prevailing taxable interest rates, but 91 percent to 98 percent of prevailing taxable interest rates.

Implicit tax should be measured by comparison of AAA municipal bonds with Treasury bonds of the same term and not by comparison with AAA corporate bonds. The market has become skeptical of the safety of AAA rating for corporations in recent years and has demanded an increasing spread between interest from AAA corpo- rate bonds vis-à-vis Treasury’s.[11] The increased risks of corporate bonds would inflate the measured implicit tax on municipals that are safer than corporate bonds. Fed- eral borrowing, moreover, is a better baseline because the federal borrowing is a possible alternative to state and local borrowing, but corporate borrowing is not. The federal government might well borrow to subsidize meritorious state and local projects, but a corporation never would.

Under a set of assumptions, including that bonds are held pro rata to overall wealth, the weighted-average tax rate of investors in municipal bonds is 31 percent.[12] On long-term borrowing, accordingly, the federal govern- ment gives up tax of approximately 31 percent of fair market taxable interest to give municipalities benefits of 2 percent to 9 percent of fair market value taxable interest. The Office of Management and Budget estimates the total cost of the municipal bond exemption to be $25.4 billion.[13] If the long-term bonds were typical of all issues, the federal government would be spending $25.4 billion to deliver a benefit to state and local governments of between $1.6 billion to $7.3 billion.[14] That ratio of cost to benefit — between 72 percent and 93 percent of the federal cost is wasted — makes section 103 a most wasteful government program.[15]

The implicit tax also creates a maximum tax on capital that a taxpayer must bear. Municipal bonds compete at the margin with all investments in the economy, and taxpayers can flee to section 103 bonds by paying a modest 2 percent to 9 percent fee. Congress must then look to other less optimal taxpayers and sources to raise the revenue it needs.

Municipal bonds also bear premium interest rates because the resale market for municipal bonds is very thin and because the market knows too little information about the various issuers.[16] It would be cheaper for the federal government to borrow directly, at its rock-bottom interest rate, and subsidize state and local governments directly because that would avoid the waste caused by illiquidity and unsatisfactory market information. In- deed, under some conditions, it would be cheaper for the municipalities to borrow on the robust market for taxable bonds because the premium arising from a thin and uninformed tax-exempt market exceeds the reduction in interest they achieve from section 103.[17]

The waste in the delivery of the benefits of municipal bonds has gotten worse over time. The amount that taxpayers are willing to pay for explicit section 103 tax exemption, by accepting lower interest rates, has de- clined over the last 10 years, as taxpayers have found more generous ways to avoid tax. Implicit taxes were just over 20 percent in 1996[18] and are now at 2 percent to 9 percent. The tax base is continually assaulted by clever tax planners and the legislative process. Taxpayers with that the other one-third of municipal bonds held by corpora- tions bears a marginal tax rate of 35 percent.

Section 146 imposes a volume cap on the amount of private activity revenue bonds that can be issued by a state, but the volume cap has been ineffective to ensure that the exemption is not swamped. An effective cap would prohibit any borrower from paying more than 65 percent of applicable federal rates for the term, and the quantity of bonds issued would drop until borrowers captured the entire forgone federal tax.

Currently, with so little of cost of the exemption being captured by borrowers, defenders of the exemption are speaking primarily for the unintended beneficiary, the middleman investors. The purported beneficiaries state and local governments are receiving only a very modest sliver of the expense.


Harm in the intended incentive. Even when state and local borrowers receive benefit from section 103, the exemption does harm. The exemption allows capital projects to go forward, which would not be justified if they had to pay the full going-interest costs.[19] The competitive market for capital filters out projects with only modest value and lets through only those uses of capital that pass over the threshold of being able to meet the prevailing interest costs. The tax exemption in section 103, however, lowers the cost of capital for some projects without requiring that they prove special merit. Capital directed to the lesser projects is pulled from better projects that would otherwise be funded.

The exemption also shifts some capital projects from private hands to public hands so that that the project can get access to the lower financing costs on municipal borrowing. The shift thus moves projects from presum- ably efficient profit-making enterprises to presumably inefficient political entities. The United States is now a net-capital-importing country borrowing marginal capital from China and other foreign sources and should not waste precious capital on lesser projects.

The game-theory situation of state and local borrowers already provides incentive for states and localities to push off public costs. Politicians trying to get elected will provide benefits to current voters, but will try to shift the costs of those benefits to the future because the future taxpayers do not vote. The tax incentive compounds the problem of shifting costs to the future. State and local governments ought to be able to borrow, but not neces- sarily at preferred rates.

The decision about which meritorious projects will receive a subsidy should be made by a competitive budget process. The federal budget is the tool by which the federal government judges the rationality of expenditures in a highly competitive environment. If a pro- gram is off budget, there is no alternative mechanism to force competition for limited resources or application of government rationality. These off-budget costs are not justified by the democratic process — because they are hidden — and not justified by a reasoned, competitive budget process.

Section 103 and other tax incentives also have no limits on the quantity issued. The uncontrolled expansion of tax-exempt and competing tax-favored investments swamp the market and make all tax incentives a wasteful mechanism.

Proposed Legislation

End of exemption and grant instead. The proposal would repeal the tax exemption for municipal bonds issued, extended, or renewed after the date of the announcement of the proposal. The proposal would replace the $24.5 billion tax expenditure cost of tax-exempt bonds with an $8 billion annual direct grant to state and local govern- ments, and to other currently eligible borrowers in pro- portion to their outstanding bonds on the announcement of the proposal. The net revenue gain from the proposal would thus be approximately $16.5 billion.

While the $8 billion grant would be to current bor- rowers pro rata to their outstanding bonds, the grants would have to be used to reduce outstanding section 103 bonds. It is inappropriate to give incentive to greater borrowing. The grants would encourage contraction of borrowing rather than expansion to overcome local in- centives to push costs off to future taxpayers who do not now vote. When no preexisting section 103 bonds remain outstanding, the grants would cease. Congress would, however, be at liberty to continue the grants at the $8 billion level or higher, with or without conditions as to their use.

Because the implicit tax on section 103 bonds is urrently so low, the $16.5 billion revenue gain can be achieved while simultaneously giving the eligible bor- rowers more benefit than they are getting from implicit tax. The implicit tax is so low that the annual grants can be increased to the point that current borrowers are active supporters of the change.

The federal government can expect to save $25.4 billion a year lost to tax exemption, less the cost of the replacement grants to current borrowers. Investors now holding tax-exempt bonds will flee to other investments with low effective tax rates. As long as the quantity of low effective tax rate investments is not allowed to expand, however, those fleeing from section 103 bonds will in turn displace other investors and send them into taxable sources.

Existing bonds. Already-issued state and local bonds can also equitably bear tax. All tax increases reduce investor income and resources. When taxes increase, however, it is more equitable to increase taxes on those who start with a windfall position of paying nothing beyond implicit tax, rather than to increase the taxes for their peers who start from paying 35 percent rates. Well-informed private investors, in general, bear risks of tax increases in the ordinary course of their business and indeed are better bearers of the risks of tax increases than is the government.[20] If the government sets a norm that taxpayers in the 35 percent tax bracket keep 65 percent of prevailing interest rates, that norm can and should be consistently maintained for all investors in the 35 percent bracket. Because current implicit taxes are modest and even insubstantial, existing investors have received windfalls, measured from the amount that they could expect from taxable bonds of like term and risk, and they have received more than necessary to induce them to buy tax-exempt bonds.

To the extent of the implicit tax, however, the federal government has received the benefit of the bargain from the section 103 grant of exemption. Thus the proposal would tax interest on existing bonds that now qualify under section 103 for interest paid in future years but would allow a credit for implicit tax borne by the investor. Both the credit and the taxable amount would be computed from the equivalent fully taxable rate on bonds of like term, risk, and date of issue. For example, assume that the interest rate on a taxable bond identical to a municipal bond in term, risk, and date of issuance was 10 percent, and the municipality had to pay 9 percent to attract investors to its like tax-exempt bonds. Ordi- narily a lender in the 35 percent tax bracket could expect to keep only 65 percent of the interest on federal or corporate bonds after tax, or here, 6.5 percent interest. After enactment of the proposal, that lender would still get 65 percent of comparable taxable interest after tax. The taxable income would be 10 percent, computed either directly from the equivalent taxable bond, or by grossing up from the municipal bond.[21] The tax would be 3.5 percent, but the 1 percent implicit tax on the bond when issued would be a credit against that 3.5 percent, so the investor would pay only 2.5 percent to the federal government. The investor would start with a 9 percent interest receipt, owe 3.5 percent tax, get a 1 percent credit for implicit tax, and would end with 6.5 percent, the same as would be achieved from the taxable bond alternative.

It should also be possible to allow a small cushion above actual implicit tax to ease administration. Thus, the taxpayer would be allowed to round up the implicit tax to the nearest one-half percent of interest. On audit and in litigation, however, the credit allowed would be the actual implicit tax borne by the holder. Eventually, the IRS should be required to publish implicit tax tables for every term, date of issuance, and risk measured by credit rating. To forgo expensive litigation, those tables would be rounded up to the nearest one-half percent interest and would be binding on all holders as to the amount of their taxable income and credit.

A state and local bond issued after the announcement date would not be entitled to credit against tax, whatever the state and local interest rate, and interest would be fully taxable to investors.

Without a tax on interest on existing bonds, holders of existing bonds can expect a windfall on top of their windfall tax-exempt interest rates, as tax-exempt bonds become rarer. Assume, for example, that bonds with an implicit tax of 5 percent of prevailing rates become so rare that purchasers are willing to take 35 percent discounts on interest — that is, the burden of normal ordinary income tax rates — to acquire the bonds. Assume a 5 percent prevailing taxable interest rate. The following chart shows how much more valuable the bond is under a 35 percent implicit tax than a 5 percent implicit tax:

Table 1. Capital Gain for Term of Years Remaining With 5% FMV Interest and 5% Implicit Tax Growing to 35%
Term Remaining (Years) Appreciation Capital Gain
30 128% 28%
25 125% 25%
20 122% 22%
15 118% 18%
10 113% 13%
5 107% 7%

If Congress were to adopt added transition relief for existing bonds (beyond the credit for implicit tax pro- posed here), the added transition relief would need to take away the windfalls such as those shown in the chart. Even for investors who do not sell, the appreciation illustrated in the chart offers them a potential to sell or borrow that is not invisible, and investors hold the bonds in the face of appreciation only because holding them is more valuable to them than realizing the appreciation. Capital gain that arises from drops of general interest rates would continue to qualify for capital gains rates, even after the proposal.

[1]Federal Reserve Bank, ‘‘Flow of Funds Accounts of the United States,’’ 89 (Table L.211) (Sept. 17, 2007), available at https://www.federalreserve.gov/releases/z1/Current/z1.pdf.



[4]Section 142(a)(1).

[5]Section 142(a)(2).

[6]Section 1400L(d)(1).

[7]Sections 55(b)(1)(A) and (B), 57(a)(5), 145.

[8]See, e.g., Senate Finance Committee Report on Revenue Bill of 1918, 69th Cong., 3d Sess. (Dec. 6, 1918) (citing ‘‘constitutional questions’’ as a reason to reject House repeal of the exemption for interest on state and local bonds).

[9]South Carolina v. Baker, 485 U.S. 505, 524, rehearing denied, 486 U.S. 1062 (1988), overruling Pollock v. Farmers’ Loan & Trust Co., 157 U.S. 429, 585-586 (1895).

[10]Using Thomson Municipal Market Monitor for Sept. 25, 2007, available at https://www.tm3.com/mmd/g_14906.html, the 30-year AAA municipal was 4.46 percent and the federal 30-year bond rate was 4.89 percent, for an implicit tax of 8.79 percent. The Wall Street Journal figures for Sept. 27, 2007, yield an implicit tax of 2.2 percent

[11]See Calvin Johnson, ‘‘A Thermometer for the Tax System: The Overall Health of the Tax System as Measured by Implicit Tax,’’ 56 S.M.U. L. Rev. 13, 38-41 (2003) (corporate bonds of increasing risk are not proper baseline for implicit tax); Cf. Aaron Lucchetti and Serena Ng, ‘‘Get Set for Wave of Debt Downgrades — With Investors Frazzled, Real-Estate Softening, Three Rating Firms Have Their Markers Out,’’ The Wall Street Journal at C1 (Nov. 9, 2007).

[12]My found weighted average marginal tax rate of holders of municipal bonds is 30.7 percent. Two-thirds of municipal bonds are held by individuals, and one-third is held by corporations. Office of Management and Budget, Fiscal Year 2008 Budget, Analytic Perspectives, available at https://www.whitehouse.gov/omb/budget/fy2008/pdf/spec.pdf (p. 294, row 158). If we assume that individually held bonds are held pro rata to capital,the found weighted average of individual holders is 28.5 percent: The found 28.5 percent assumes that the top 1 percent by wealth is in the 35 percent tax bracket; the next 4 percent are in the 33 percent bracket; the next 15 percent are in the 25 percent tax bracket; the next 25 percent are in the 15 percent bracket; and the bottom half are in the 10 percent or 0 percent bracket, which are fair but rough approximations. Income distributions are from the President’s Advisory Panel on Federal Tax Reform, ‘‘Simple, Fair, and Pro-Growth: Proposals to Fix

America’s Tax System’’ at 30 (Nov. 2005) and wealth percentages are from Marco Cagetti and Mariacristina De Nardi, ‘‘Wealth Inequality: Data and Model,’’ Federal Reserve Bank of Chicago

Working Paper 2005-10, at 4 (Aug. 17, 2005). The assumption is

[13]OMB, supra note 12, p. 294, row 158 (sum of $8 billion for corporations and $17.4 billion for individuals).

[14]If 31 percent of tax-exempt interest is $24.5 billion, tax- exempt interest is $82 billion. Then 2 percent of $82 billion is $1.6 billion, and 9 percent of $82 billion is $7.3 billion.

[15]That distinction as ‘‘most wasteful’’ is, however, shared by other programs artificially reducing the definition of income to give incentives. All the exemption and deduction incentives share the same market for tax benefits.

[16]Chunchi Wu, Yan He, Haitao Li, and Junbo Wang, ‘‘Liquid- many alternatives are not willing to take, and do not need to take, very much reduction in interest to get fully legal section 103 tax exemption.


[18]Calvin Johnson, supra note 12, at 23 (finding implicit tax of 20.20 percent in 1996)

[19]Cf. M. Mussa and R. Kormendi, The Taxation of Municipal Bonds 189-192 (American Enterprise Institute, 1979) (opposing increased delivery-efficiency for municipal bonds because it would reduce overall economic efficiency).

[20]See, e.g., Michael J. Graetz, ‘‘Legal Transitions: The Case of Retroactivity in Legal Revision,’’ 126 U. Penn. L. Rev. 47 (1977).

[21]Grossing up is the process familiar to tax specialists of going from an after-tax amount to the gross receipt taxable amount. For an implicit tax of it, gross-up would find the taxable interest by dividing the municipal bond interest by (1-it).

Previously published by the University of Texas – School of Law, December 2007

To top