Why are municipal bonds tax-exempt?
(This is part three of a series on infrastructure investment.)
As I explained in my previous post, the municipal bond market is the primary vehicle for infrastructure investment in the United States. Because most municipal bonds are issued on a tax-exempt basis, changes to the federal tax code have the potential to disrupt infrastructure investment significantly.
This exemption is frequently and foolishly a target for federal lawmakers on both the right and the left, however, who are either penny-pinching or looking for funding to offset their political agendas.
Open versus closed markets for securities
The municipal bond market functions as a mostly closed market. Both the supply side (entities that borrow money through the issuance of bonds) and the demand side (bondholders) are restricted by tax policy. This is very different than the market for corporate bonds, for example, which involves diverse participants.
Only state and local governments, nonprofits, and corporations that have received an allocation of a state’s private activity bond cap (discussed later) may borrow money on a tax-exempt basis.
Similarly, demand for municipal securities is limited to investors that can benefit from the tax exemption. This essentially eliminates pensions and foreign investors. However, pensions and foreign investors can and do still invest in domestic infrastructure projects through other vehicles, such as private equity funds.
(For an excellent long-form discussion on this topic, I would recommend George Friedlander of Citigroup’s The Structure of Demand for Tax-Exempt Securities: Past, Present, and Future.)
So tax policy keeps the market for the bonds that finance infrastructure projects mostly closed and it could theoretically be changed to open the market to new investors. Why should tax policy remain unchanged?
This meta-structure persists for good reasons: (1) however counter-intuitive this may seem, the benefits associated with a closed municipal bond market outweigh the costs, and (2) alternative tax structures have proven problematic. As I will explain in future posts, Congress does not have to alter tax policy to “open up” the market for infrastructure projects. Direct private investment can have the same net effect.
Tax reform and municipal bonds
Tax reform proposals often focus on eliminating or radically transforming the federal government’s tax expenditures. In what follows, I explain (1) how some popular attitudes toward tax reform are based on a facile understanding of tax expenditures, and (2) why there is not a market for the substitutes for tax-exempt municipal bonds that lawmakers have proposed.
What is a tax expenditure?
Before launching into a discussion of tax expenditures, it would probably be useful to explain what tax expenditures are. The Congressional Budget and Impoundment Control Act of 1974 defines tax expenditures as “revenue losses attributable to provisions of federal tax laws which allow a special exclusion, exemption, or deduction from gross income or which provide a special credit, a preferential rate of tax, or deferral of tax liability.”
These provisions are included in tax law either to (1) provide an incentive for specific behaviors or (2) reduce an economic burden for specific people. Among the largest tax expenditures, for example, are the exclusion of employer health insurance and pensions; mortgage interest deduction; Medicare exclusion; capital gains rates; earned income credit; deduction of charitable contributions; and exemption for interest income on municipal bonds.
Direct expenditures and tax expenditures are simple policy alternatives from the perspective of lawmakers. As Linda Sugin explains in Tax Expenditures, Reform, and Distributive Justice:
They are called “tax expenditures” because they achieve the same objectives as direct federal spending, but they are administered through tax law. They are unlike “real” tax provisions that are necessary to accurately measure taxpayer income; without them, the tax law would still carry out its revenue collection function. Only the mechanism for administration distinguishes tax expenditures from direct expenditures — instead of the government allocating funds for particular programs, tax expenditures allow taxpayers to reduce their tax payments by participating in various activities.
Arguments for eliminating all or most tax expenditures
Advocates for tax reform usually have two objectives in common: (1) to simplify the tax code, which can be achieved by either broadening the tax base or lowering rates, and (2) to reduce the federal deficit. Tax expenditures represent a convenient target for achieving these objectives. In recent years, many proposals predicated on eliminating tax expenditures have been promoted by the White House and members of Congress.
These proposals are popular for a number of reasons, both philosophical and practical. A sweeping elimination of tax expenditures would remove provisions that reform proponents regard as “extraneous to the fundamental purpose of tax law,” as Sugin puts it, and condense those labyrinthine tax forms every constituent despises.
Reformers also argue that eliminating tax expenditures could produce a significant amount of new revenue necessary to bring the federal government’s fiscal situation back into balance or to offset lower tax rates.
Finally, some reformers would suggest that a large fraction of these revenues are lost to people who could stand to pay a lot more in taxes and that eliminating “tax loopholes for the rich” is a moral imperative. This is especially true for the exemption for municipal bond interest, whereby the tax benefits are distributed regressively.
Problems with this worldview
The problem with this fixation on tax expenditures, and with most political discourse these days, is that it caricatures complicated and important issues. It is completely appropriate to frame tax expenditures as a form of spending rather than taxing when measuring their impact on the federal budget. It is absurdly bad policy, however, to treat tax expenditures monolithically when gauging their social legitimacy, for no other reason than the fact that they share the same administrative mechanism (or, to use plain English, because they all happen to be included in the federal tax code).
Some tax expenditures play a central role in policy, and the exemption for interest earned on municipal bonds — which dates back over a century to the institution of the federal income tax in 1913 — is one of them. From a policy standpoint, the tax exemption for municipal bonds serves two purposes.
The tax exemption offsets the cost of essential government services
First, the tax exemption for municipal bonds is one component of the federal government’s financial contribution to construct and maintain infrastructure across the country, and I think everyone would agree that this is an essential government service. Yes, the federal government forgoes collecting revenue on municipal debt, but the exemption lowers the cost of capital projects (roads, bridges, water and sewer treatment plants and transmission lines, schools, university buildings, and so on) for state and local governments. This happens because investors are (under ordinary conditions) willing to accept a lower interest rate on the bonds in consideration of the tax benefits they receive from holding them.
It is short-sighted to argue that only wealthy investors benefit from the municipal bond exemption because they are the ones who have large incomes to shelter from taxation. Entire populations (and multiple generations) benefit from using the projects constructed with the proceeds of municipal bond issues. This is, in fact, the whole point of a subsidy in the first place — to provide an incentive for those with resources to use them in a way that benefits society.
If the federal government eliminates the tax benefits for holding municipal bonds, it will become more expensive for state and local governments to undertake new capital projects as interest rates normalize, meaning that either fewer projects will be feasible or infrastructure investment will crowd out other political priorities. Eliminating a “tax loophole for the rich” will produce very real social costs for everyone else.
Anyone who has tracked the progress of federal highway legislation recognizes the importance of preserving the tax exemption as a resource for infrastructure investment instead of moving exclusively to direct appropriations. The federal government has relied on short-term extensions for transportation spending since 2009, when the last long-term highway plan expired, so that lawmakers could bicker over the Keystone XL pipeline, environmental issues, and spending in general. Lawmakers have also failed to take any action on the perennially insolvent Highway Trust Fund, which has required several transfers from the general fund.
Local access to capital is essential to federalism
Second, the tax exemption for municipal bonds provides market access for smaller governmental entities by giving investors a reason to purchase their bonds rather than the bonds of large corporations or any other asset. This means that a school district can issue bonds to construct a school and the cost of that school will be borne by the taxpayers in the immediate vicinity who actually benefit from the facility.
If the tens of thousands of small borrowers in the municipal bond market lost market access, their only alternative would be to consolidate governmental entities — and, by extension, decision-making about revenue collection and services. Fiscal conservatives who prefer small government rather than an expansive and expensive bureaucracy should think comprehensively about the consequences of eliminating the tax exemption for municipal bonds. This includes whether they want to live in a world where towns have to beg the state for money when they need to construct a fire station and whether they want that fire station to compete with other state priorities (e.g., pension contributions) for funding.
It should also be noted that, from a credit perspective, the decentralized nature of the municipal bond market is one of its main strengths. This is a feature that limits contagion and contributes to our nation’s financial stability. Furthermore, in the wake of Detroit settling its general obligation bonds at far less than face value, investors’ preference for a specific revenue pledge has grown stronger. Removing the tax-exemption would force consolidation, which is the exact opposite of demand dynamics in the marketplace.
The bottom line here is that the tax exemption for municipal bonds deserves to be considered within the context of how to pay for extending and improving infrastructure in this country over the long term and how to divide responsibility for those expenses among various stakeholders. This is a slightly more important conversation to have than whether the tax code should reward people for giving their old clothes to Goodwill, but lawmakers are treating these concerns as if they are equivalent when they talk about tax expenditures as an undifferentiated mass. If that sounds reckless, it is because it is reckless.
The "incentivizing borrowing" straw man
Some federal policymakers have argued that the tax exemption for municipal bonds provides an incentive for state and local governments to take on more debt than they would if they had to borrow at higher taxable rates. It is factually incorrect to say lower rates contribute to increased borrowing at the state and local levels.
The post-recession municipal market has demonstrated this point quite clearly. While interest rates have remained at historic lows, the municipal market has experienced net negative issuance. This means that the volume of borrowing for new projects has been so low that it has been eclipsed by the debt that has been refinanced or retired. The market has actually shrunk considerably. Conversely, this year the issuance of bonds by corporations hit a record $1.5 trillion. So corporations are taking advantage of a low interest rate environment, but governments are not.
What is the reason for this phenomenon? Unlike the federal government, most state and local governments have to adopt balanced budgets and pay down existing debt. This means that other political priorities can outweigh undertaking new capital projects even when interest rates are low. Post-recession, state and local governments have considered restoring funding to programs that had been cut significantly a higher priority than capital projects.
Exploiting low interest rates also requires agility that governments often do not have. It takes months to years to conceptualize, plan, design, and then ultimately borrow money to construct new infrastructure. A government cannot borrow money with a vague anticipation of constructing something. (Also, arbitrage provisions in the federal tax code effectively prohibit this. Not to mention the fact that bondholders do not just give money to governments and say “go forth and do good things.” They tend to want their money attached to an identified project.)
Puerto Rico is probably the only issuer in the municipal market that can be said to have had an incentive to borrow more due to favorable tax policy. Puerto Rico has the distinction of being triple exempt. For this reason, the debt of Puerto Rico and its agencies has been uniquely desirable to mutual funds and other investors historically. This has provided an artificial source of demand that has allowed the island to borrow billions of dollars even as it approaches insolvency.
It is also worth noting that the federal tax code imposes a number of limitations on how tax-exempt bond proceeds may be utilized. Some of the more controversial forms of borrowing, such as pension obligation bonds, already must be issued on a taxable basis.
Whatever one thinks of government spending, removing the exemption for municipal bonds is an illogical target.
The private activity bond straw man
I have noticed some commentators citing the use of private activity bonds as a way the tax-exemption has been abused and offering this as a reason the tax exemption should be eliminated. This reflects a fundamental ignorance about both private activity bonds and the municipal market as whole.
Let’s begin with a brief discussion about what private activity bonds (PABs) are. PABs are tax-exempt bonds that governments issue on behalf of private borrowers because these private projects can be said to serve a public purpose. Since the Tax Reform Act of 1986, states have been subject to aprivate activity bond volume cap. Essentially, the limit of tax-exempt PABs that can be issued within a state is determined by multiplying a per capita amount (that is set annually by the Internal Revenue Service and adjusted for inflation) by the state’s population. The issuance of private activity bonds represents a minor fraction of the municipal bond market.
I would concede that occasionally PABs are used to finance some truly jackass projects (ask Curt Schilling), especially in the context of economic development. This is largely a factor of how individual states choose to allocate their volume cap to potential borrowers policy-wise. For the most part, however, these bonds are used to finance projects that do serve public purposes: housing revenue bonds that benefit low-to-moderate income and first-time homebuyers; affordable multifamily housing; projects undertaken by nonprofits; educational facilities; hospitals; redevelopment of blighted areas and so on. (Nonprofits do not need an allocation of volume cap but must pass more stringent private use tests, among other forms of scrutiny.) PAB issues are subject to TEFRA approvals, which include a public hearing to receive input from the community.
(NB: Most of the PABs for transportation projects are part of a separate authorization.)
It is bizarre that anyone would treat a minor fraction of a minor fraction of a multi-trillion-dollar marketplace as an excuse to alter tax policy in toto or be oblivious to the public goods that are served by this provision.
Problems with the federal government's revenue loss estimates
For many years, economists have challenged the methodology used by the Treasury Department and JCT in estimating the amount of revenue the federal government loses on the exemption for interest on municipal bonds and, consequently, the amount of revenue the government stands to gain from repealing the provision. If you ever want to raise the blood pressure of economists and bond analysts, bring this topic up at an industry conference.
Their ire is well-founded. According to James Poterba and Arturo Ramirez Verdugo in their paper, Portfolio Substitution and the Revenue Cost of Exempting State and Local Government Interest Payments from Federal Income Tax, Treasury and JCT assume that if individual investors did not hold tax-exempt bonds, they would simply hold taxable bonds instead. Treasury and JCT are essentially suggesting that investors would not change their behavior if they lost this particular way of sheltering income from taxation. (Yes, I know, I could probably just stop here…)
As Poterba and Verdugo point out, the federal government is likely to pick up less revenue than Treasury and JCT estimate:
In all likelihood, more complex portfolio adjustments would follow changes in the tax treatment of interest payments on state and local government bonds. Taxable bonds are among the most heavily taxed portfolio assets, so assuming that the high-marginal-tax rate household who owns tax-exempt bonds would shift to holding highly-taxed taxable bonds if the tax-exemption were repealed is likely to overstate the revenue cost of the tax exemption. If repeal of the tax exemption leads current holders of tax-exempt bonds to substitute into other lightly taxed assets, such as common stocks with low dividend yields, then the revenue gain from repeal could be much smaller than calculations based on the taxable bond substitution suggest.
In a discussion on this topic a while back, my friend @groditi mentioned some other good reasons to expect investor behavior not to reflect JCT’s assumptions. While a change in the tax status of the bonds would not change the credit component of a bond, the duration component would be affected due to the inherent rates sensitivity of tax-exempt securities. In a rising rate environment, investors would demand a higher premium to account for the increased price volatility of taxable bonds.
Poterba and Verdugo examined a variety of plausible portfolio substitution strategies and concluded that the revenue cost of the exemption for municipal bonds “might be as little as half the standard estimate.” Would it seem worthwhile to lawmakers to disrupt the $3.5 trillion municipal bond market just to pick up a few billion dollars? Is there nothing else that the federal government can cut besides infrastructure investment?
(Note: Poterba provided additional information regarding his analysis at a recent conference.)
In my next installment, I will talk about shadow demand in the municipal bond market and why the Build America Bond program won’t be returning anytime soon.
