Reexamining the Tax Exemption of Municipal Bond Interest
- Since the enactment of the federal income tax in 1913, interest on state and local bonds has been excluded from taxation. However, the original reason for this exclusion – concern about the constitutionality of taxing the borrowing power of state and local governments – is likely no longer applicable.
- The strongest economic justification for the tax exemption of municipal bonds is that it encourages state and local governments to invest in infrastructure projects that create benefits for nonresidents. On the other hand, there is also reason to believe that the tax exemption will cause municipalities to overinvest in infrastructure, particularly if states and localities are also able to shift their tax burdens onto nonresidents.
- A tax exclusion is an unideal policy design for subsidizing state and local debt: it delivers larger benefits for taxpayers in higher income brackets, shuts some investors completely out of the municipal bond market, and makes the subsidy difficult for Congress and voters to evaluate.
- Most importantly, there is a compelling case that the current tax treatment of municipal bond interest is inefficient. For every dollar that the federal government forgoes due to the provision, state and local governments receive less than a dollar in lower borrowing costs; the remainder goes largely to high-income households.
Under the current U.S. tax code, individuals and corporations that own bonds are generally required to pay taxes on their interest income. However, since the enactment of the federal income tax in 1913, interest on state and local bonds has been tax-exempt.
Over the next ten years, the federal government will forgo as much as $617 billion in revenue by excluding interest on state and local bonds from income taxation. As a result, the exclusion of municipal bond interest is one of the largest tax expenditures in the individual income tax code.
In this paper, I argue that lawmakers should seriously consider limiting, reforming, or eliminating the exclusion of municipal bond interest. While the federal government may have a legitimate role in subsidizing state and local government spending, the current exclusion of municipal bond interest has several flaws and drawbacks, which I describe at length.
History of the Tax Exemption of Municipal Bond Interest
For more than 200 years, state and local governments have issued bonds to fund capital investments. The first municipal bond was issued in 1812, by New York City, to fund a canal. Throughout the 1800s, state and local governments were the primary investors in public infrastructure, building railroads, canals, highways, water and sewer systems, school buildings, and other improvements. During this time period, the municipal bond market grew rapidly.
The question of whether the federal government should tax interest on municipal bonds was first raised following the passage of the Wilson-Gorman Tariff Act of 1894. This bill established the first peacetime national income tax, a levy of 2 percent on all income over $4,000. Importantly, the new income tax also applied to interest income from state and local bonds.
Less than a year later, the U.S. Supreme Court ruled the new income tax unconstitutional, in the famous case of Pollock v. Farmers’ Loan & Trust Co. (1895). The primary reason for the decision was the Court’s finding that the new tax violated Article I, Section 2 of the Constitution, which requires that all direct taxes be apportioned among the states according to their population. However, the Court also touched on the issue of municipal bonds, arguing that a federal tax on state and local bond interest would be unconstitutional.
Regarding municipal bonds, the Court held that levying a federal tax on state and local bond interest would violate the constitutional doctrine of intergovernmental tax immunity: the principle that federal government cannot impose a tax on income derived from the activities of a state. Explaining this logic, Chief Justice Melville Fuller wrote, “The tax in question is a tax on the power of the States and their instrumentalities to borrow money, and [is] consequently repugnant to the Constitution.”
Less than two decades later, the Sixteenth Amendment was ratified, allowing the federal government to levy income taxes on “incomes, from whatever source derived.” While the plain text of the Amendment seems to allow for federal taxes on income from municipal bonds, proponents of the Amendment argued that this was not part of its purpose. For instance, Sen. William Borah (R-ID) commented, in 1910, “To construe the proposed amendment so as to enable us (the Congress) to tax the instrumentalities of the state would do violence to the rules laid down by the Supreme Court for a hundred years [and] wrench the whole institution from its harmonious proportions…”
Following the passage of the Sixteenth Amendment, the Revenue Act of 1913 enacted the current income tax and specifically excluded municipal bond interest from taxation. Almost immediately, this provision attracted controversy from those who believed that it gave an unfair borrowing advantage to state and local governments. The Coolidge, Harding, and Hoover administrations all supported amending the Constitution to explicitly eliminate the tax exemption of municipal bond interest. However, for most of the 20th century, the tax treatment of municipal bonds was left unchanged.
A turning point came in the early 1980s, when Congress passed several laws in succession that curtailed the favorable tax treatment of municipal bonds. The Tax Equity and Fiscal Responsibility Act of 1982 (TEFRA) required states and localities to issue their bonds in registered form in order for the bonds to be tax-exempt. The Social Security Amendments of 1983 modified the taxation of Social Security benefits such that some households with high municipal bond income were subject to additional taxes on their benefits. Finally, the Tax Reform Act of 1986 created several restrictions on arbitrage bonds (where a municipality reinvests the proceeds of tax-exempt bonds in other investments with higher yields) and private-activity bonds (where a municipality issues bonds to fund private projects).
These bills caused the Supreme Court to revisit the question of whether federal taxes on municipal bonds are constitutional in South Carolina v. Baker. In a 7-1 decision, the Court upheld Congress’ restrictions on tax-exempt municipal bonds, arguing that “the rationale underlying Pollock… has been thoroughly repudiated.” In essence, the Court ruled that taxing the interest received by someone who owns a municipal bond is different than taxing a state or local government, and that the former is constitutionally permissible.
As a result of South Carolina v. Baker, the original rationale for the tax exemption of municipal bond interest no longer applies. As a result, proponents of the municipal bond interest exclusion now tend to justify the continued existence of the provision in terms of its role in subsidizing infrastructure investment.
Over the past few decades, there have been several calls for the tax exemption on municipal bonds to be limited or repealed. For instance, in 2010, the Simpson-Bowles Commission on Fiscal Responsibility and Reform called for eliminating the tax exemption on all interest from new municipal bonds. Overall, one Wall Street Journal reporter estimates that, since 1918, there have been “no fewer than 125” proposals to limit or eliminate the exclusion of municipal bond interest.
Tax-Exempt Municipal Bonds Today
About 6 percent of all bonds issued each year in the United States are municipal bonds, almost all of which are tax-exempt. According to the IRS, state and local governments issued $421 billion in tax-exempt bonds in 2013.
Tax-exempt bonds issued by state and local governments fall into one of two categories: governmental bonds or private-activity bonds. The majority of tax-exempt bonds issued each year are governmental bonds, which means that they are issued for public purposes, guaranteed by public funds, or both. Governmental bonds are typically used to fund public educational facilities, transportation infrastructure, utilities, and other improvements.
On the other hand, a bond is categorized as a private-activity bond if more than 10 percent of the proceeds of a bond are dedicated for private business use and more than 10 percent of the principal is secured by a private business. Only certain types of private-activity bonds are tax-exempt, and their issuance is subject to several rules and regulations. Private-activity bonds are typically used to fund hospitals, private educational facilities, housing, airports, and other improvements.
Out of the $421 billion of tax-exempt bonds that state and local governments issued in 2013, $340 billion were governmental bonds (80.6 percent) and $81 billion were private-activity bonds (19.4 percent). This ratio is fairly typical: between 1991 and 2013, governmental bonds accounted for 75.3 percent of all tax-exempt bonds, while the other 24.7 percent were private-activity bonds.
The majority of municipal bonds are issued by the state and local governments of just a few states. In 2013, issuers in California, New York, Texas, Pennsylvania, Illinois, Ohio, and New Jersey accounted for more than half of new bond issuances.
Municipal bonds are owned disproportionately by households, rather than banks and other financial institutions, compared to other types of bonds: 42.9 percent of all municipal debt was held by the household sector in 2015. By contrast, in the same year, the household sector only held 11.5 percent of all debt in the U.S. economy. This is because many financial institutions are unable to take advantage of the tax-exemption of municipal bonds.
A Framework for Evaluating the Tax Exemption of Municipal Bond Interest
The debate over the tax exemption of municipal bond interest is sometimes dominated by anecdotes and narrow arguments. Opponents of the exclusion tend to draw attention to the most prominent abuses of the current system, such as the use of tax-exempt municipal bonds to fund local stadiums and arenas. Meanwhile, proponents of the exclusion often highlight particularly attractive state and local infrastructure projects that are subsidized by tax-exempt bonds, without considering larger policy questions.
It would be more productive for the policy community to consider large structural questions about the exclusion of municipal bond interest than to dwell on the specific merits of the thousands of state and local projects funded by the provision. As a starting point for this discussion, the following framework may be useful for evaluating the exclusion of municipal bond interest:
- Should the federal government subsidize state and local borrowing (and, by extension, state and local infrastructure spending)?
- If so, should this subsidy take the form of a tax exclusion?
- If so, is the current treatment of municipal bond interest efficient and equitable?
Broadly speaking, these questions point to the best direction forward for the tax treatment of municipal bond interest. If the answer to all three questions is “yes,” then Congress should preserve the tax exemption of municipal bond interest in its current form. If the answer to the third question is “no,” then Congress ought to alter or limit the exclusion to better accomplish its purpose. If the answer to the last two questions is “no,” then Congress ought to convert the tax exclusion into a tax credit or direct grant. Finally, if the answer to all three questions is “no, then Congress should repeal the exclusion of municipal bond interest in full.
The next three sections of this paper will address these questions.
Justifying a Federal Subsidy for State and Local Borrowing
The most basic question that pertains to the tax treatment of municipal bonds is whether the federal government should be subsidizing state and local borrowing (or infrastructure spending) in the first place.
There is a compelling theoretical case to be made that such a subsidy will lead state and local governments to spend too much on infrastructure. For instance, we can imagine a state government deciding whether to spend $10 million on building a new highway, which is expected to deliver $9 million in economic benefits. Because the costs of the highway outweigh the expected benefits, then the state government should not construct it. However, if the state receives a $1.5 million subsidy from the federal government for building a new highway, it will go ahead with the project, even though the highway is a socially wasteful investment.
However, there is also a case that, in the absence of a subsidy for infrastructure spending, state and local governments would spend too little on infrastructure. Here, we can imagine a state government deciding whether to spend $10 million on a new highway, which is expected to deliver $11 million in economic benefits. However, $2 million of the benefits of the highway will go to individuals and businesses outside of the state, such that the state government expects the highway project to deliver only $9 million in benefits to its residents. If the state only cares about helping its own residents, it will not go forward with the project. However, if the state receives a $1.5 million subsidy from the federal government for building a highway, it will go ahead with the project; the federal subsidy will incentivize the state to make a socially beneficial investment.
This is the standard economic argument for a federal subsidy for state and local infrastructure spending: without such a subsidy, state and local governments might underinvest in infrastructure projects that benefit nonresidents.
As a result, the desirability of a federal subsidy for state and local infrastructure spending depends on two empirical questions. First, to what extent do state and local infrastructure projects actually benefit nonresidents? Second, to what extent are state and local governments already able to shift their tax burden to nonresidents, without the help of a federal subsidy?
The first question – regarding the extent to which the benefits of state and local infrastructure spill over to nonresidents – is extremely difficult to answer. Writing in 2001, Mila Freire and Richard E. Stren comment, “The basic problem… is that no one, anywhere, has a good idea of the magnitude of spillovers associated with particular services.” It is not even entirely clear which categories of state and local infrastructure have the largest spillovers to nonresidents. For instance, it might seem intuitive that state highways would deliver significant benefits to out-of-state residents, but a well-known 1995 paper by Douglas Holtz-Eakin and Amy Ellen Schwartz found no evidence that state highways deliver productivity benefits beyond a state’s borders.
Because it is difficult to assess whether the benefits of an infrastructure project will spill over to nonresidents, it is likely that any federal subsidy for state and local infrastructure spending will end up subsidizing some projects that only benefit residents. In these cases, the federal government would be encouraging state and local governments to overinvest in infrastructure – an inevitable side effect of subsidizing infrastructure spending.
The second question – regarding the extent to which state and local governments are able to shift their tax burdens onto nonresidents – is a crucial one. To return to the example above: if a state government were considering a $10 million highway that will deliver $9 million in benefits to its residents and $2 million in benefits to nonresidents, it would ordinarily not construct the highway. However, if the state is able to raise an additional $2 million in taxes on out-of-state commuters, then it will indeed go ahead with the project. In this case, there would be no need for a federal subsidy to incentivize the state to make a socially beneficial investment.
Indeed, there is evidence that states and localities are already able to shift their tax burdens to nonresidents, without the help of federal subsidies – a practice known as “tax exporting.” A recent Tax Foundation report estimates that 22 percent of state and local taxes are collected from nonresidents, through sales taxes on tourists, income taxes on commuters, and similar measures.
The fact that state and local governments have the ability to collect taxes from nonresidents undercuts the standard economic case for a federal subsidy for state and local investment. In the absence of such a federal subsidy, states and localities may still have sufficient incentives to fund socially beneficial projects that benefit nonresidents, because of the ability to export their tax burdens.
To sum up, federal policymakers should be wary of the possibility that subsidizing state and local investments will lead these governments to spend too much on infrastructure. At the same time, they should be eager to find ways to target federal subsidies towards those state and local investments with the largest spillover benefits.
Choosing the Best Policy Design
Moving to a more concrete question, it is worthwhile to consider whether a federal subsidy for state and local infrastructure should be administered as a tax exclusion. Before evaluating this question, however, it is necessary to describe exactly how the tax exclusion of municipal bond interest works, and why it leads to subsidized borrowing costs for state and local government.
The basic theory behind the tax exclusion of municipal bond interest is that investors are willing to accept lower interest payments on tax-free bonds. As a result, state and local governments will be able to pay lower interest rates on their debt and still be able to attract investors. This leads to a lower cost of borrowing for states and localities, allowing them to invest more in infrastructure.
To take an example, if households face a marginal tax rate of 39.6 percent, they will be indifferent between a corporate bond that pays a 3 percent interest rate and an equally risky municipal bond that pays a 1.812 percent interest rate (3 percent × (1 – 0.396)). As a result, state and local governments will only have to pay an interest rate of 1.812 percent on their debt in order to stay competitive in the bond market, rather than the going rate of 3 percent.
By its very nature, a tax exclusion of municipal bond interest will lead to several policy design problems. For instance, with a tax exclusion, the total size of the federal subsidy for state and local borrowing is tied to the individual income tax rate schedule. For example, when Congress lowered the top individual tax rate in 2001 from 39.6 percent to 35 percent, it also decreased the size of the federal subsidy for municipal debt, for reasons entirely unconnected to the condition of the nation’s infrastructure or state and local finances. Needless to say, this is an unideal way to determine the size of a federal subsidy.
Similarly, a tax exclusion for municipal bond interest will always deliver smaller benefits to households in lower tax brackets. A household in the 39.6 tax bracket may be willing to purchase a municipal bond with a 1.812 percent interest rate, but a household in the 25 percent bracket would see a higher after-tax return from the corporate bond with a 3 percent rate.
In fact, a tax exclusion for municipal bond interest delivers no benefits to investors who do not pay taxes, such as foreign investors and pension funds. This means that these investors are effectively shut out of the municipal bond market, unwilling to purchase municipal bonds with reduced interest rates. Some analysts have claimed that this feature of the market makes municipal bonds relatively opaque, with little oversight from institutional investors.
There are other important concerns with a tax exclusion serving as a vehicle for a federal subsidy for state and local infrastructure. For instance, tax expenditures – provisions of the tax code that resemble spending programs – are often not transparent. Many voters may be unaware of the size and scope of the federal subsidy for state and local investment because it is tucked away in the tax code, in the form of an exclusion.
Finally, using a tax exclusion to subsidize state and local infrastructure may make it especially difficult for Congress to exercise oversight over the subsidy. The federal appropriations process is an important opportunity for Congress to evaluate the effectiveness of different spending programs and to prioritize the most important ones. On the other hand, most provisions in the tax code – even those that resemble spending programs – do not receive regular oversight, and occur largely on autopilot.
As a result of all of these concerns, many economists and policymakers over the years have proposed turning the tax exclusion for municipal bond interest into a tax credit or into a direct grant for state and local governments. Most of these proposals would be significant improvements over the current tax exclusion: they would provide a uniform benefit to bondholders of all income levels; they would encourage institutional investors to take part in the municipal bond market; and they would allow voters and Congress to more easily evaluate the subsidy.
An Inefficient Subsidy to State and Local Governments
There is a strong case that the current tax treatment of municipal bond interest is inefficient: for every dollar of revenue that the federal government forgoes because of the exclusion of municipal bond interest, state and local governments receive less than a dollar in lower borrowing costs.
Why is this the case? We can return to the example from the last section, of a household in the 39.6 percent bracket that is indifferent between a corporate bond that pays a 3 percent interest rate an and equally risky tax-exempt municipal bond that pays a 1.812 percent interest rate. In theory, this should allow states and localities to issue bonds with an interest rate of 1.812 percent.
However, state or local governments often want to attract investors who are not in the top marginal income tax bracket. As a result, they often need to issue municipal bonds with higher interest rates. For instance, if a state wants to attract a household in the 25 percent bracket, it will need to offer an interest rate of 2.25 percent (3 percent × (1 – 0.25)). As a result, even though a household in the top bracket would be willing to buy a municipal bond with an interest rate of 1.812 percent, in this scenario it would be able to buy one at the prevailing rate of 2.25 percent.
If the state pays the high-income household $100 in interest, the household will see tax savings of $39.60 (the $100 of interest × the 39.6 percent bracket). However, the state will only save $25.00 in borrowing costs (the $100 of interest × the 25 percent reduction in the interest rate). In other words, the federal government will lose $39.60 in revenue in order to provide a $25.00 subsidy to the municipality.
This is the most important criticism of the current tax treatment of municipal bond interest: as a rule, the federal government will lose more because of the provisions than state and local government will receive. In other words, high-income households are able to capture a portion of the benefits from the municipal bond interest exclusion.
While the interest rates in the examples above are imaginary, this phenomenon has been observed empirically in municipal bond markets time and again. In 2009, the Congressional Budget Office suggested that only 80 percent of the tax expenditure for tax-exempt municipal bonds translates into lower state and local borrowing costs; the other 20 percent of the tax expenditure goes to high-income households. Using a different methodology, in 2007, University of Texas law professor Calvin H. Johnson estimated that between 72 and 93 percent of the tax expenditure for municipal bonds is captured by high-income households, while only 7 to 28 percent goes to state and local governments.
Two Competing Cases of Tax Neutrality
Up to this point, this paper has discussed the tax exemption of municipal bond interest using terms such as “subsidy” – as if it were just like any federal spending program. However, it is also useful to evaluate the treatment of municipal bond interest as a tax provision, as part of the larger question, “Should interest income be generally excluded from taxes?”
There is a case to be made that the individual income tax should not apply to interest income at all. If the principal of a bond investment has already been subject to taxes, then taxing the interest income would result in a double tax on household saving.
To take an example, suppose that a household makes $100 in wages, pays $20 in taxes on its earnings, and invests the remaining $80 in a bond with a 5 percent yield. The household’s interest income on the bond has already been lowered due to taxes; instead of earning $5 of interest on a $100 investment, it will only earn $4 of interest on an $80 investment. Taxing the household’s interest income would result in a double tax, and would encourage the household to consume immediately rather than saving.
To the extent that all interest income should be tax-exempt, the exclusion of municipal bond interest could make sense, as a step toward a broader effort to make the tax code neutral between consumption and saving. However, it is also the case that the current exclusion of municipal bond interest is highly non-neutral: it disproportionately benefits a narrow category of investment.
Ultimately, lawmakers who care about the concept of tax neutrality will have to decide whether it is more significant that the exclusion for municipal bond interest helps make the tax code more neutral between consumption and saving, or whether it is more significant that the exclusion helps make the tax code less neutral between different investments.
Like almost every provision in the tax code, the tax exemption for municipal bond interest has many passionate defenders. One recent report went so far as to argue that “Congress and the Administration need to positively assert the tax-exempt expenditure is sacrosanct.”
In this paper, I have tried to highlight some of the shortcomings of the current tax treatment of municipal bond interest. While intended as a subsidy for state and local infrastructure spending that benefits nonresidents, the exclusion of municipal bond interest also applies to many state and local spending projects that do not need to be subsidized. The provision is designed poorly, delivering larger benefits for taxpayers in higher income brackets, and shutting some investors completely out of the municipal bond market. Finally, the subsidy is inefficient, as only a portion of every dollar forgone by the federal government ends up in the hands of state and local governments.
This evidence points to a clear takeaway: no provision in the U.S. tax code should be “sacrosanct.” The U.S. tax code is badly in need of reform, and any tax reform effort will need to limit or eliminate tax expenditures in order to broaden the federal tax base. As Congress moves forward with discussions about tax reform, it should keep every tax provision on the table, including the exclusion of municipal bond interest.
 This paper uses “state and local bonds” and “municipal bonds” interchangeably. In addition, “exclusion” and “exemption” are used interchangeably.
 “Tax Expenditures,” U.S. Department of the Treasury, Nov. 11, 2015, https://www.treasury.gov/resource-center/tax-policy/Documents/Tax-Expenditures-FY2017.pdf. This is a static estimate, which assumes that taxpayers would not shift their investments from debt to equity if the exclusion were repealed. See: James Poterba and Arturo Ramírez Verdugo, “Portfolio Substitution and the Revenue Cost of the Federal Income Tax Exemption for State and Local Government Bonds,” National Tax Journal 64 (2011): 591-614.
 “From New York Canals to High-Tech Infrastructure: the History of Municipal Bonds,” Neighborly, https://learn.neighborly.com/municipal-bond-guide/history-of-municipal-bonds-chapter-2/
 Mayraj Fahim, “Municipal bonds have been issued by US local government since 1812,” City Mayors, March 2012, http://www.citymayors.com/finance/bonds.html
 John Joseph Wallis, “American Government Finance in the Long Run: 1790 to 1990,” The Journal of Economic Perspectives 14 (2000): 61-82.
 Joe Henchman, “Today in History: Income Tax Ruled Unconstitutional in Pollock v. Farmers Loan Trust Co.,” Tax Foundation, April 2013, https://taxfoundation.org/blog/today-history-income-tax-ruled-unconstitutional-pollock-v-farmers-loan-trust-co
 For background on intergovernmental tax immunity, see Carter Glass, “A Review Of Intergovernmental Immunities From Taxation,” Washington and Lee Law Review 4 (1946): 48
 Pollock v. Farmers’ Loan & Trust Co. 157 U.S. 429 (1895). Accessed at https://supreme.justia.com/cases/federal/us/157/429/case.html
 The Constitution of the United States, Amendment 16.
 “Income Tax a Necessity,” The Herald Democrat, Feb. 11, 1910, https://www.coloradohistoricnewspapers.org/cgi-bin/colorado?a=d&d=THD19100211-01.2.10#
 Revenue Act of 1913, Section II, B.
 Peter Fortune, “The Municipal Bond Market, Part I: Politics, Taxes, and Yields,” New England Economic Review, 1991, https://www.bostonfed.org/economic/neer/neer1991/neer591b.pdf
 Tax Equity and Fiscal Responsibility Act of 1982, Pub. L. 97-248, Sec. 215.
 Social Security Amendments of 1983, Pub. L. 98-21, Sec. 121(a). As a result, it is not correct to claim that municipal bonds are entirely tax-exempt under the current federal tax code. A household that receives more municipal bond interest in a given year could face higher income taxes as a result, because of the formula for calculating the tax on Social Security benefits.
 Tax Reform Act of 1986, Pub. L. 99-514, Sec. 1301(a).
 South Carolina v. Baker 485 U.S. 505. Accessed at http://caselaw.findlaw.com/us-supreme-court/485/505.html
 James E. Spiotto has argued that South Carolina v. Baker “may not be the last word” on the constitutionality of taxing municipal bond interest, although his analysis seems to be based less on an assessment of current precedent than speculation that a future Court might roll back the ruling in South Carolina. See “The Renewed Battle over Tax Exemption of Interest on State and Local Government Debt Obligations,” Government Finance Review, 2013, http://www.chapmanstrategicadvisors.com/media/publication/2_Renewed_Battle_
 Andrew Ackerman, “Hundreds of Local Officials Defend Municipal-Bond Tax Exemption,” The Wall Street Journal, 2016, http://www.wsj.com/articles/hundreds-of-local-officials-defend-municipal-bond-tax-exemption-1456830002
 “The Moment of Truth,” The National Commission on Fiscal Responsibility and Reform, 2010, https://www.fiscalcommission.gov/sites/fiscalcommission.gov/files/documents/
 Jason Zweig, “How Long Will the Tax Break on Municipal Bonds Last?” The Wall Street Journal, 2011, http://www.wsj.com/articles/SB10001424052748704810504576307233579693982
 “Statistics,” Securities Industry and Financial Markets Association, http://www.sifma.org/research/statistics.aspx
 “Tax Exempt Bond Statistics,” IRS, https://www.irs.gov/uac/soi-tax-stats-tax-exempt-bond-statistics
 “Table 2. Long-Term Tax-Exempt Governmental Bonds, by Bond Purpose and Type of Issue, 2013,” IRS, https://www.irs.gov/uac/soi-tax-stats-tax-exempt-bond-statistics
 26 U.S. Code § 141(a)
 “Tax Expenditures,” U.S. Department of the Treasury, 2016, https://www.treasury.gov/resource-center/tax-policy/Documents/Tax-Expenditures-FY2017.pdf
 “Tax Exempt Bond Statistics,” IRS, https://www.irs.gov/uac/soi-tax-stats-tax-exempt-bond-statistics
 Ibid. This average omits the year 2005, for which the IRS has not published figures.
 “Financial Accounts of the United States: L.212 Municipal Securities,” Federal Reserve, as of June 9, 2016, https://www.federalreserve.gov/releases/z1/current/
 “Financial Accounts of the United States: L.208 Debt Securities,” Federal Reserve, as of June 9, 2016, https://www.federalreserve.gov/releases/z1/current/
 See, for example, Eliot Brown, “Use of Taxpayer Money for Pro-Sports Arenas Draws Fresh Scrutiny,” The Wall Street Journal, 2015, http://www.wsj.com/articles/use-of-taxpayer-money-for-pro-sports-arenas-draws-fresh-scrutiny-1425856677
 See, for example, “The Impacts of Altering Tax-Exempt Municipal Bond Financing on Public Drinking Water & Wastewater Systems,” Association of Metropolitan Water Agencies, 2013, http://www.amwa.net/sites/default/files/AMWA-NACWA_MuniBondAnalysis_July13.pdf
 Here, “too much” refers to a scenario where a government purchases an additional unit of infrastructure, even though the costs of the project are larger than the societal benefits it will bring. Similarly, “too little” refers to a scenario where a government declines to purchase an additional unit of infrastructure, even though its societal benefits would outweigh its costs. Of course, cost-benefit analysis should not be the final word on questions of public policy, but it is useful for this sort of theoretical evaluation.
 Robert H. Gordon, “An Optimal Taxation Approach to Fiscal Federalism,” National Bureau of Economic Research, Working Paper No. 1004 (1982), http://www.nber.org/papers/w1004.pdf
 This argument is expressed well in a recent Congressional Research Service report: “Some public goods and services are best provided by state or local governments. Some of the goods and services provided by state or local governments, however, benefit both residents, who pay local taxes, and nonresidents, who pay minimal if any local taxes. Since state and local taxpayers are likely to be unwilling to provide these services to nonresidents without compensation, it is probable that state and local services will be underprovided. In theory, the cost reduction provided by the exemption of interest income compensates state and local taxpayers for benefits provided to nonresidents. This encourages the governments to provide the optimal amount of public services.” Steven Maguire and Jeffrey Stupak, “Tax-Exempt Bonds: A Description of State and Local Government Debt,” Congressional Research Service, 2015, https://www.fas.org/sgp/crs/misc/RL30638.pdf
 Mila Freire and Richard Stren, The Challenge of Urban Government: Policies and Practices, Washington: World Bank Publications, 2001.
 Douglas Holtz-Eakin and Amy Ellen Schwartz, “Spatial Productivity Spillovers from Public Infrastructure: Evidence from State Highways,” National Bureau of Economic Research, Working Paper No. 5004 (1995), http://www.nber.org/papers/w5004.pdf
 “State-Local Tax Burden Rankings FY 2012,” Tax Foundation, 2016, https://taxfoundation.org/article/state-local-tax-burden-rankings-fy-2012
 One additional consideration relevant to this point is that many state and local governments have historically high levels of debt. The last few years have seen several high-profile municipal defaults, including Stockton, Calif., Detroit, and Puerto Rico. To the extent that the federal tax exemption of municipal bonds encourages state and local governments to borrow more, it leads to a higher risk of municipal default. Regarding the specific relationship between tax-exempt bonds and Puerto Rico’s default, see Scott Greenberg and Gavin Ekins, “Tax Policy Helped Create Puerto Rico’s Fiscal Crisis,” Tax Foundation, 2015, https://taxfoundation.org/blog/tax-policy-helped-create-puerto-rico-s-fiscal-crisis
 Indeed, municipal bond interest rates are lower than corporate bond interest rates for this very reason.
 Edward V. Regan, “A New Approach to Tax-Exempt Bonds,” The Jerome Levy Economics Institute of Bard College, 1999, http://www.levyinstitute.org/pubs/ppb58.pdf
 Whitney Afonso, “The Challenge of Transparency in Taxation,” Mercatus Center, 2015, http://mercatus.org/publication/challenge-transparency-taxation
 See, for example: Calvin H. Johnson, “Repeal Tax Exemption for Municipal Bonds,” Tax Notes, 2007, https://law.utexas.edu/faculty/calvinjohnson/repeal-exempt-muni-bonds.pdf; The Bipartisan Tax Fairness and Simplification Act of 2011, S. 727, 112th Congress, https://www.gpo.gov/fdsys/pkg/BILLS-112s727is/pdf/BILLS-112s727is.pdf; Marc Joffe, “The Smartest Way to Finally Fix America’s Crumbling Infrastructure,” The Fiscal Times, 2016, http://www.thefiscaltimes.com/Columns/2016/04/12/Smartest-Way-Finally-Fix-America-s-Crumbling-Infrastructure; “Subsidizing Infrastructure Investment with Tax-Preferred Bonds,” Congressional Budget Office, 2009, https://www.cbo.gov/sites/default/files/111th-congress-2009-2010/reports/10-26-taxpreferredbonds.pdf
 For a more detailed explanation of this criticism, see Peter Fortune, “The Municipal Bond Market, Part II: Problems and Policies,” New England Economic Review, 1992, https://www.bostonfed.org/economic/neer/neer1992/neer392d.pdf
 “Subsidizing Infrastructure Investment with Tax-Preferred Bonds,” Congressional Budget Office, 2009, https://www.cbo.gov/sites/default/files/111th-congress-2009-2010/reports/10-26-taxpreferredbonds.pdf
 Calvin H. Johnson, “Repeal Tax Exemption for Municipal Bonds,” Tax Notes, 2007, https://law.utexas.edu/faculty/calvinjohnson/repeal-exempt-muni-bonds.pdf
 Ronald P. Bernardi et. al., “Tax Exempt Municipal Bonds: The Case for an Efficient, Low-Cost, Job-Creating Tax Expenditure,” Bernardi Securities Inc., 2011, http://www.munibondsforamerica.org/cms/wp-content/uploads/2012/11/Bernardi-Securities-Inc.-White-Paper.pdf