Tax-Exempt Bonds: A Description of State and Local Government Debt

Tax-Exempt Bonds: A Description of
State and Local Government Debt
Updated March 26, 2008
Steven Maguire
Specialist in Public Finance
Government and Finance Division

Tax-Exempt Bonds: A Description of State and Local
Government Debt
This report provides information about state and local government debt. State
and local governments often issue debt instruments in exchange for the use of
individuals’ and businesses’ savings. This debt obligates state and local governments
to make interest payments for the use of these savings and to repay, at some time in
the future, the amount borrowed. State and local governments finance capital
facilities with debt rather than out of current tax revenue in order to match the time
pattern of benefits from these capital facilities with the time pattern of tax payments.
The federal government subsidizes the cost of most state and local debt by
excluding the interest income from federal income taxation. This tax exemption of
interest income is granted because it is believed that state and local capital facilities
will be under provided if state and local taxpayers have to pay the full cost.
State and local debt is issued as bonds, to be repaid over a period of time greater
than one year and perhaps exceeding 20 years, and as notes, to be repaid within one
year. General obligation bonds are secured by the promise to repay with general tax
revenue, and revenue bonds are secured with the promise to use the stream of
revenue generated by the facility built with the bond proceeds. Most debt is issued
to finance new capital facilities, but some is issued to refund a prior bond issue
(usually to take advantage of lower interest rates). Tax-exempt bonds issued for
some activities are classified as governmental bonds and can be issued without
federal constraint because most of the benefits from the capital facilities are enjoyed
by the general public. Many tax-exempt revenue bonds are issued for activities
Congress has classified as private because most of the benefits from the activities
appear to be enjoyed by private individuals and businesses. The annual volume of
a subset of these tax-exempt private-activity bonds is capped.
Arbitrage bonds devote a substantial share of the proceeds to the purchase of
assets with higher interest rates than that being paid on the tax-exempt bonds. Such
arbitrage bonds are not tax exempt because Congress does not want state and local
governments to issue tax-exempt bonds and use the proceeds to earn arbitrage profits.
The arbitrage profits could substitute for state and local taxes.
The major policy issue in this area is the effort to use tax-exempt bonds to
increase federal financial support for a variety of public facilities. Another policy
issue is whether constraints should be relaxed on the types of activities for which
state and local governments can issue tax-exempt debt. The extent to which the
current arbitrage bond rules prohibit what some consider legitimate state and local
financial behavior is a related area of dispute. The list of activities that classify tax-
exempt private-activity bonds — and whether they should be included in the volume
cap — is another area of controversy. This report will be updated as new data
become available.

What is Debt?.....................................................1
Why Do State and Local Governments Issue Debt?.......................1
What Makes State and Local Debt Special?.............................2
What Does Tax Exemption Cost the Federal Government?.................4
Why Does the Federal Government Subsidize State and Local Debt?.........5
Classifying State and Local Debt Instruments............................5
Maturity: Short-Term vs. Long-Term..............................5
Tax and Revenue Anticipation Notes..........................5
Auction Rate Securities.....................................6
Security: General Obligation, Revenue, and Lease Rental Bonds........7
Use of the Proceeds: New-issue vs. Refunding Bonds.................9
Public Purpose vs. Private Purpose...............................10
Private Activities Eligible for Tax Exemption.......................10
What Are Arbitrage Bonds?.........................................14
What Are Tax Credit Bonds?........................................15
Legislative Issues.................................................17
Suggested Readings...............................................18
List of Tables
Table 1. Yield on Tax-Exempt and Corporate Bonds of Equivalent Risk,
the Yield Spread, and the Yield Ratio:1980 to 2007...................3
Table 2. Tax Expenditure on the Outstanding Stock of Public Purpose
Tax-Exempt Bonds: 1994 to 2007.................................4
Table 3. Volume of State and Local Tax-Exempt Debt: 1980 to 2007........6
Table 4. Volume of Long-Term Tax-Exempt Debt: General Obligation
(GO), Revenue, and Refunding Bonds, 1980 to 2007..................8
Table 5. Private-Activity Bond Volume by Type of Activity in
2005 and 2006...............................................10
Table 6. New-Money, Long-Term Private-Activity Bond Volume as
Percent of Total Bond Volume, 1988 to 2005.......................11
Table 7. Federal Tax Expenditure for Selected Private Activities Financed
with Tax-Exempt Bonds.......................................12
Table 8. Comparison of Three Types of Bonds with a 6% After-Tax Yield...17

Tax-Exempt Bonds: A Description of State
and Local Government Debt
What is Debt?
Individuals and businesses lend their accumulated savings to borrowers. In
exchange, borrowers give lenders a debt instrument. These debt instruments,
typically called bonds, represent a promise by borrowers to pay interest income to
lenders on the principal (the amount of money borrowed) until the principal is repaid
to the lenders. This principal, sometimes called the proceeds, generally is used to
finance the construction of capital facilities.
Why Do State and Local Governments Issue Debt?
Since public capital facilities provide services over a long period of time, it
makes financial and economic sense to pay for the facilities over a similarly long
period of time. This is particularly true for state and local governments. Their
taxpayers lay claim to the benefits from these facilities by dint of residency and
relinquish their claim to benefits when they move. Given the demands a market-
oriented society places on labor mobility, taxpayers are reluctant to pay today for
state and local capital services to be received in the future. The rational response of
the state or local official concerned with satisfying the preferences of constituents is
to match the timing of the payments to the flow of services, precisely the function
served by long-term bond financing. An attempt to pay for capital facilities “up
front” is likely to result in a less than optimal rate of public capital formation.
State and local governments are also faced with the necessity of planning their
budget for the year (or in some cases for two years). This requires a balancing of
revenue forecasts against forecasts of the demand for services and spending. Not
infrequently, the inevitable unforeseen circumstances that undermine any forecast
cause a revenue shortfall, which must be financed with short-term borrowing, or
“notes.” In addition, even when the forecasts are met, the timing of expenditures
may precede the arrival of revenues, creating the necessity to borrow within an
otherwise balanced fiscal year. Finally, temporarily high interest rates that prevail
at the time bonds are issued to finance a capital project may induce short-term
borrowing in anticipation of a drop in rates.
Thus, state and local governments have valid reasons to borrow funds. In fact,
these reasons are so universally accepted that both taxpayers and the courts have

ignored the nineteenth century legacy of unrealistically restrictive constitutional and
statutory limitations on state and local borrowing.1
What Makes State and Local Debt Special?
The federal government has chosen to intervene in the public capital market by
granting the debt instruments of state and local governments a unique privilege —
the exemption of interest income earned on these bonds from federal income tax.
The tax exemption lowers the cost of capital for state and local governments, which
should then induce an increase in state and local capital formation. The lower cost
of capital arises because investors would be indifferent between taxable bonds (e.g.,
corporate bonds) that yield a 10% rate of return before taxes and tax-exempt bonds
of equivalent risk that yield a 6.5% return. The taxable bond interest earnings carry
a tax liability (35% of the interest income in most cases) making the after tax return
on the two bonds identical at 6.5%. Thus, state and local governments could raise
capital from investors at an interest cost 3.5 percentage points (350 basis points)
lower than a borrower issuing taxable debt.
Generally, the degree to which tax-exempt debt is favored is measured in two
ways. The yield spread is the difference between interest rate on corporate bonds and
the interest rate on municipal bonds of equivalent risk. Table 1 lists the average
yield on high-grade municipal bonds and AAA-rated corporate bonds from 1980 to
2007, and the corresponding yield spread. The spread grew to a high of 3.43% in
1980 and dropped to a low of 0.94% in 2003. The greater the yield spread, the
greater is the nominal savings to state and local governments as measured by the
interest rates they would have to pay if they financed with taxable debt. As the
spread approaches zero, state and local borrowing costs approach the level of taxable
bond interest rates.
Another measure, the yield ratio (which is an average rate on tax-exempt bonds
divided by an average rate on a corporate bond of like term and risk), adjusts the
spread for the level of interest rates. The lower the ratio, the greater the savings to
state and local governments relative to taxable debt. As the ratio approaches one, the
cost of tax-exempt state and local borrowing approaches that of taxable borrowing.
As shown in Table 1, the ratio was lowest in 1980 at 0.71 and reached a peak of 0.84
in 1982. For the last decade the ratio has been relatively stable.
These variations in the cost of state and local borrowing relative to the cost of
taxable borrowing depend upon the demand for and supply of both tax-exempt and
taxable bonds. Demand for tax-exempt bonds depends upon the number of investors,
their wealth, statutory tax rates, and alternative investment opportunities. Supply
depends upon the desire of the state and local sector for capital facilities and their
ability to engage in conduit financing (issuing state or local government bonds and
passing the proceeds through to businesses or individuals for their private use).

1 Dennis Zimmerman, “History of Municipal Bonds,” in his The Private Use of Tax-Exempt
Bonds: Controlling the Public Subsidy of Private Activity (Washington, The Urban Institute
Press, 1991), pp. 17-27.

Almost all of the factors which influence demand and supply are affected by federal
tax policy.
Table 1. Yield on Tax-Exempt and Corporate Bonds of
Equivalent Risk, the Yield Spread, and the Yield Ratio:
1980 to 2007
High Grade
Tax-ExemptAAA CorporateYield Ratio
YieldYieldYield Spread(tax-exempt/
Year (%) (%) (%) corporate)
1980 8.51 11.94 3.43 0.71
1981 11.23 14.17 2.94 0.79
1982 11.57 13.79 2.22 0.84
1983 9.47 12.04 2.57 0.79
1984 10.15 12.71 2.56 0.80
1985 9.18 11.37 2.19 0.81
1986 7.38 9.02 1.64 0.82
1987 7.73 9.38 1.65 0.82
1988 7.76 9.71 1.95 0.80
1989 7.24 9.26 2.02 0.78
1990 7.25 9.32 2.07 0.78
1991 6.89 8.77 1.88 0.79
1992 6.41 8.14 1.73 0.79
1993 5.63 7.22 1.59 0.78
1994 6.19 7.96 1.77 0.78
1995 5.95 7.59 1.64 0.78
1996 5.75 7.37 1.62 0.78
1997 5.55 7.26 1.71 0.76
1998 5.12 6.53 1.41 0.78
1999 5.43 7.04 1.61 0.77
2000 5.77 7.62 1.85 0.76
2001 5.19 7.08 1.89 0.73
2002 5.05 6.49 1.44 0.78
2003 4.73 5.67 0.94 0.83
2004 4.63 5.63 1.00 0.82
2005 4.29 5.24 0.95 0.82
2006 4.42 5.59 1.17 0.79
2007 4.42 5.52 1.10 0.80
Source: Council of Economic Advisors, Economic Report of the President, February 2008, Table

What Does Tax Exemption Cost the
Federal Government?
The direct cost to the federal government of this interest exclusion is the
individual and corporate income tax revenue forgone. Consider a 35% marginal tax
rate corporate investor who purchases a 6.5% tax-exempt bond with principal of
$1,000 that is to be repaid after 20 years. Each year for 20 years this taxpayer
receives $65 in tax-exempt interest income. Each year the federal government
forgoes collecting $35 of revenue because the revenue loss is based upon the yield
the taxpayer forgoes. For example, if the investor had purchased a taxable bond
carrying a 10% interest rate, he would have received $100 in interest income and paid
$35 in income taxes on that income.2
The annual federal revenue loss (or tax expenditure) on the outstanding stock
of tax-exempt bonds is reported in the Analytical Perspectives section of the Budget
every year. The estimates for the last 14 years are displayed in Table 2.3 Because
they are based upon the outstanding stock of public-purpose tax-exempt bonds, it
takes time for some legislative changes to show up in these data. The amount of
forgone tax revenue from the exclusion of interest income on public-purpose tax-
exempt bonds is substantial, $23.5 billion in 2007.
Table 2. Tax Expenditure on the Outstanding Stock of
Public Purpose Tax-Exempt Bonds: 1994 to 2007
(in billions)
YearTax ExpenditureYearTax Expenditure
1994 $19.6 2001 $27.4
1995 $20.4 2002 $29.9
1996 $24.9 2003 $31.1
1997 $19.9 2004 $26.2
1998 $24.6 2005 $26.4
1999 $27.5 2006 $23.0
2000 $26.8 2007 $23.5
Source: Office of Management and Budget. Analytical Perspectives: Budget of the United States
Government, various years.

2 The decision about preferred alternatives is critical to estimates of the revenue loss from
tax-exempt bonds. An entire range of financial and real assets exists with different yields,
risk, and degree of preferential taxation. It is not true that the municipal bond purchaser’s
preferred alternative is always a taxable bond.
3 These estimates are derived by summing the revenue loss estimates for each activity listed
in the tax expenditures budget. Technically, this is incorrect because each activity’s revenue
loss is calculated in isolation, and there are interactive effects. Nonetheless, without an
estimate of the interactive effects’ impact on revenue loss, the summing employed here
provides the best available order of magnitude.

Why Does the Federal Government Subsidize
State and Local Debt?
When first introduced in 1913, the federal income tax excluded the interest
income earned by holders of the debt obligations of states and their political
subdivisions from taxable income. It was asserted by many that any taxation of this
interest income would be unconstitutional because the exemption was protected by
the Tenth Amendment and the doctrine of intergovernmental tax immunity. The U.S.
Supreme Court rejected this claim of constitutional protection in 1988 in South
Carolina v. Baker (485 U.S. 505, [1988]).
Although the legal basis for the subsidy is statutory rather than constitutional,
the policy reason for the subsidy is economic. Economic theory suggests that certain
types of goods and services will not be provided in the correct or “optimal” amounts
by the private sector because some of the benefits are consumed collectively, a street
light for example. The Nation’s welfare can be increased by public provision of
these goods and services, and some of these public goods and services are best
provided by state or local governments. However, some of the goods and services
provided by state or local governments benefit both residents, who pay taxes, and
nonresidents, who pay minimal if any 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 under provided. In
theory, the cost reduction provided by the exemption of interest income compensates
state and local taxpayers for benefits provided to nonresidents and encourages these
governments to provide the optimal amount of public services.
Classifying State and Local Debt Instruments
State and local debt can be classified based on (1) the maturity (or term), which
is the length of time before the principal is repaid; (2) the type of security, which is
the financial backing for the debt; (3) the use of the proceeds for either new facilities
or to refinance previously-issued bonds; and (4) whether the type of activity being
financed has a public or a private purpose. Another important factor is the level risk.
Just about every bond issued by a state or local government is rated based on the
probability of default. The privately managed rating agencies incorporate all of the
above factors as well as the financial health of the entity issuing the bonds when
arriving upon a bond rating. The higher the default risk, the lower the rating.
Maturity: Short-Term vs. Long-Term
Tax and Revenue Anticipation Notes. State and local governments must
borrow money for long periods of time and for short periods of time. Long-term debt
instruments are usually referred to as bonds, and carry maturities in excess of one
year. Short-term debt instruments are usually referred to as notes, and carry
maturities of 12 months or less. If the notes are to be paid from specific taxes due
in the near future, they usually are called tax anticipation notes (TANs); if from
anticipated intergovernmental revenue, they are called revenue anticipation notes
(RANs). If the notes are to be paid from long-term borrowing (e.g., bonds), they are

called bond anticipation notes (BANs). Tax anticipation notes and revenue
anticipation notes are often grouped together and referred to as tax and revenue
anticipation notes (TRANs). Table 3 displays the volume of long-term and short-
term borrowing since 1980. Long-term borrowing dominates state and local debt
activity in most years, with the long-term share peaking in 1985 at over 90% of this
Table 3. Volume of State and Local Tax-Exempt Debt:
1980 to 2007
YearShort-Term(in millions)Long-Term(in millions)Long-Term Share of Total

1980 $26,485 $47,133 64.0%

1981 $34,443 $46,134 57.3%

1982 $43,390 $77,179 64.0%

1983 $35,849 $83,348 69.9%

1984 $31,068 $101,882 76.6%

1985 $20,809 $206,991 90.9%

1986 $22,046 $150,638 87.2%

1987 $20,518 $105,027 83.7%

1988 $23,666 $117,316 83.2%

1989 $29,596 $125,005 80.9%

1990 $34,804 $127,828 78.6%

1991 $44,800 $172,443 79.4%

1992 $42,894 $234,667 84.5%

1993 $47,354 $292,249 86.1%

1994 $40,293 $165,034 80.4%

1995 $38,346 $159,983 80.7%

1996 $41,695 $185,014 81.6%

1997 $46,434 $220,672 82.6%

1998 $34,584 $286,817 89.2%

1999 $36,511 $227,741 86.2%

2000 $41,249 $200,880 83.0%

2001 $56,610 $288,083 83.6%

2002 $72,386 $358,569 83.2%

2003 $69,771 $383,498 84.6%

2004 $56,951 $359,695 86.3%

2005 $50,544 $408,193 89.0%

2006 $44,126 $388,656 89.8%

2007 $57,999 $429,727 88.1%

Source: The Bond Buyer Yearbook, 2007 and earlier editions.
Auction Rate Securities. Auction Rate Securities (ARSs) are long-term
debt obligations with the unique feature of adjustable or variable interest rates. In
contrast to long-term, fixed rate securities, issuers go to auction periodically
(anywhere from every 7 to 35 days) to reset the interest rate on the debt outstanding.
The auction mechanism and interest rate parameters vary by issuer (and issue) though
most use what is termed a “Dutch auction” where each bidder submits a bid for the
amount they are willing to purchase at a given interest rate. All bids are ordered from
lowest interest rate to highest interest rate and the rate where the market clears, e.g.,

where all bonds would be purchased, establishes the new ARS rate. All bidders
receive that rate.
Unique Features of ARSs. ARSs typically have a “call option” where the
issuer can buy the ARS back at par (face value) at any scheduled auction and then
retire the debt. Most ARSs are insured by the issuer because they do not carry a
“put” option that would allow bondholders to sell the bonds at a specified price to the
issuer or a designated third party. The bond insurance reduces risk and thus interest
rate making the bonds less costly to issuers. For this reason, ARSs are
very sensitive to changes in credit ratings and normally require the highest
ratings (e.g. AAA/Aaa) to make them marketable. This is usually achieved with4
bond insurance.
A Failed Auction. The existing holders of ARSs offer bids as well as new
bidders. If all bids of both existing bond holders and new participants fail to clear
the market, the auction is termed a “failed auction.” In this scenario, the original
agreement with bondholder stipulates a “reservation” interest rate the issuer must pay
in the event of a failed auction at least until the next successful auction. The
reservation rate is typically significantly higher than current market interest rates.
Because the rate is higher than market interest rates, issuers of ARSs wish to avoid
paying the reservation rate.
Growth of ARSs. The issuance of ARSs has grown considerably in the last

20 years. In 1988, the Bond Buyer identified one ARS issue valued at $25 million;5

none were issued in 1987. In 2004, the peak year, 438 ARS bonds valued at $42.5
billion were issued. The number of ARSs issues has since dropped to 322 and $38.7
billion in 2007.
Security: General Obligation, Revenue,
and Lease Rental Bonds
Another important characteristic of tax-exempt bonds is the security provided
to the bondholder. General obligation (GO) bonds pledge the full faith and credit of
the issuing government. The issuing government makes an unconditional pledge to
use its powers of taxation to honor its liability for interest and principal repayment.
Revenue bonds, or non-guaranteed debt, pledge only the earnings from revenue-
producing activities, most often the earnings from the facilities being financed with
the revenue bonds. Should these earnings prove to be inadequate to honor these
commitments, the issuing government is under no obligation to use its taxing powers
to finance the shortfall. Some revenue bonds are issued with credit enhancements
provided by insurance or bank letters of credit that guarantee payment upon such a
revenue shortfall.

4 Douglas Skarr, California Debt and Investment Advisory Commission, “Auction Rate
Securities,” Issue Brief, Aug. 2004, pp. 2-3.
5 The Bond Buyer 2007 Yearbook, SourceMedia Inc., New York, NY.

The first two columns of Table 4 display the breakdown between long-term GO
and revenue bonds since 1980. The long-term market has been and continues to be
dominated by revenue bonds, which are nonguaranteed debt instruments. During the

1960s (not shown in this table), revenue bonds constituted less than 40% of long-

term bond volume; during the 1970s (also not shown) the revenue bond share crept
into the 50% to 65% range; and beginning in 1979 this share settled into the high 60s
and low 70s, achieving a high of 73.1% in 1988. The revenue bond share has
remained at or just below 70% from 1989 through 2007.
Table 4. Volume of Long-Term Tax-Exempt Debt: General
Obligation (GO), Revenue, and Refunding Bonds, 1980 to 2007
(in millions)
Long-Term Bond VolumeRefunding Bonds
YearGeneral RevenueRevenueAmountRefunding
Obligation Share Share

1980 $16,347 $30,786 65.3% $1,649 3.5%

1981 $13,988 $32,146 69.7% $1,192 2.6%

1982 $23,276 $53,903 69.8% $4,044 5.2%

1983 $22,584 $60,764 72.9% $13,048 15.7%

1984 $27,508 $74,374 73.0% $11,390 11.2%

1985 $55,287 $148,994 72.9% $57,867 28.3%

1986 $45,555 $105,417 69.8% $56,063 37.1%

1987 $30,867 $74,656 70.7% $38,490 36.5%

1988 $31,502 $85,509 73.1% $36,591 31.3%

1989 $38,501 $86,504 69.2% $28,842 23.1%

1990 $40,303 $87,526 68.5% $19,881 15.6%

1991 $57,110 $115,334 66.9% $41,444 24.0%

1992 $80,479 $154,188 65.7% $92,446 39.4%

1993 $91,555 $200,694 68.7% $150,152 51.4%

1994 $55,767 $109,267 66.2% $38,601 23.4%

1995 $60,367 $99,615 62.3% $33,850 21.2%

1996 $64,344 $120,670 65.2% $45,941 24.8%

1997 $72,297 $148,201 67.2% $60,161 27.3%

1998 $93,644 $193,031 67.3% $81,957 28.6%

1999 $71,046 $156,695 68.8% $38,330 16.8%

2000 $66,573 $134,308 66.9% $19,587 9.8%

2001 $101,652 $186,431 64.7% $64,699 22.5%

2002 $125,675 $232,894 65.0% $92,350 25.8%

2003 $142,767 $240,731 62.8% $95,240 24.8%

2004 $129,527 $230,168 64.0% $88,276 24.5%

2005 $144,170 $264,023 64.7% $130,877 32.1%

2006 $114,858 $273,798 70.4% $79,163 20.4%

2007 $131,676 $298,051 69.4% $75,654 17.6%

Source: The Bond Buyer Yearbook, 2007 and earlier editions.
All tax-exempt interest income attributable to state and local governments does
not appear in the form of bonds. Governments may engage in installment purchase
contracts and finance leases for which the portion of the installment or lease payment
to a vendor is tax exempt. For example, computer equipment or road building
equipment could be leased from a vendor using a rental agreement or an installment

sales contract. Under this type of agreement, the monthly payments to the vendor are
large enough to cover the vendor’s interest expense on the funds borrowed to
purchase the equipment which was leased to the government. This portion that is
attributable to interest income is not included in the vendors taxable income. Such
transactions are often referred to as municipal leasing.
Lease rental revenue bonds and certificates are variations on revenue bonds. An
authority or nonprofit corporation issues bonds, builds a facility with the proceeds,
and leases the facility to a municipality. Security for the bonds or certificates is
based on the lease payments. When the bonds are retired, the facility belongs to the
lessee (the municipality). An advantage to this type of arrangement is that many
states’ constitutional and statutory definitions do not consider this type of financing
to be debt because the lease payments are annual operating expenses based upon
appropriated monies.
The leasing technique has also been used to provide tax-exempt funds to
nonprofit organizations. A municipality issues the bonds for the construction of a
facility that is leased to a nonprofit hospital or university. Again, security for the
bonds is based on the lease payments.
Use of the Proceeds: New-issue vs. Refunding Bonds
Long-term tax-exempt bond issues also can be characterized by their status as
new issues or refunding issues. New issues represent bonds issued to finance new
capital facilities. Refundings usually are made to replace outstanding bonds with
bonds that carry lower interest rates or other favorable terms. As such, the refunding
bonds usually do not add to the stock of outstanding bonds or the capital stock. The
proceeds of the refunding bonds are used to pay off the remaining principal of the
original bond issue, which is retired. Advance refunding bonds, however, do add to
the outstanding stock of bonds without adding to the stock of capital. Advance
refunding bonds are issued prior to the date on which the original bonds are refunded,
so that for a period of time there are two bond issues outstanding to finance the same
capital facilities.
The last two columns of Table 4 show the dollar value of refunding issues and
their share of total long-term bond volume. The share varies widely, depending to
a great extent on changes in the relative magnitudes of taxable and tax-exempt
interest rates. Note that the 1993 increase in the top marginal individual income tax
rates may have increased the demand for tax-exempt bonds. Higher tax rates make
tax-exempt bonds more attractive relative to taxable bonds, all other things being
equal. The increased demand and accompanying lower interest rates may have
prompted state and local governments to replace outstanding issues with refunding
bonds that carried lower interest rates. In contrast, refundings dropped considerably
in 1999 and 2000. The decline could have been in response to higher interest rates
or to strong economic conditions in most states which minimized the need for debt
finance generally. The story is reversed from 2001 to 2003 as the economy slowed
and state budgets were strained by lower tax revenue collections. New issues in 2003
were more than double the amount of new issues in 2000. In 2005, GO bonds and
refunding bond volume peaked, likely reflecting the historically low interest rates on
tax-exempt debt (see Table 1).

Public Purpose vs. Private Purpose
An important characteristic of tax-exempt bonds is the purpose or activity for
which the bonds are issued. Most of the tax legislation pertaining to tax-exempt
bonds over the last 30 years reflects an effort to restrict tax exemption to bonds
issued for activities that satisfy some broadly defined “public” purpose, that is, for
which federal taxpayers are likely to receive substantial benefits. Bonds are
considered to be for a public purpose if they satisfy either of two criteria: less than
10% of the proceeds are used directly or indirectly by a non-governmental entity; or
less than 10% of the bond proceeds are secured directly or indirectly by property used
in a trade or business. Bonds that satisfy either of these tests are termed
“governmental” bonds and can be issued without federal limit. Bonds that fail both
of these tests are termed “private-activity” bonds (PABs) because they provide
significant benefits to private individuals or businesses. These projects are ineligible
for tax-exempt financing.
Activities which fail the two tests but are considered to provide both public and
private benefits have been termed eligible or qualified PABs. These selected
activities can be financed with tax-exempt bonds. Table 5 provides the dollar value
of new issues of tax-exempt private-activity bonds and their share of total private-
activity volume capacity for 2005 and 2006. Table 6 provides historical data on the
portion of PAB volume to total bond volume.
Table 5. Private-Activity Bond Volume by
Type of Activity in 2005 and 2006
Issued inPortion of Available
Private Activity(in millions)Capacity
2005 2006 2005 2006
Total Volume Capacity Available$49,142$48,675100.0%100.0%
New Volume Capacity$26,079$26,43853.1%54.3%
Carryforward from Previous Years$23,063$23,27746.9%47.8%
Carryforward to Next Year$26,337$22,63853.6%46.5%
Single-family Mortgage Revenue $6,507$10,09313.2%20.7%
Multi-family Housing$5,562$6,25211.3%12.8%
Student Loans$5,124$4,01810.4%8.3%
Exempt Facilities$1,915$2,6053.9%5.4%
Abandon Capacity$910$1,4171.9%2.9%
Industrial Development$1,000$1,1952.0%2.5%
Housing not Classified$822$5621.7%1.2%
Mortgage Credit Certificates$493$5101.0%1.0%
Other Activities$485$2841.0%0.6%
Source: State Allocations of Private-Activity Bonds in 2005,” The Bond Buyer, May 1, 2006; and
State Allocations and Use of Private-Activity Bonds in 2006,” The Bond Buyer, June 25, 2007.
Private Activities Eligible for Tax Exemption
All tax-exempt private-activity bonds are subject to restrictions that do not apply
to governmental bonds, chief among them being no advance refundings and the

inclusion of the interest income in the alternative minimum income tax base. In
addition, the annual dollar value of all bonds issued for most of these activities by all
governmental units within a state is limited to the greater of $85 per resident or
$262.095 million in 2008. The cap has been adjusted for inflation since 2004. The
annual volume cap applies to the total of bonds issued primarily for but not limited
to multi- and single-family housing, industrial development, exempt facilities,6
student loans, and bond-financed takeovers of investor owned utilities (usually
electric utilities).
Table 6. New-Money, Long-Term Private-Activity Bond Volume
as Percent of Total Bond Volume, 1988 to 2005
YearTotal Bond Volume(in millions)Private Activity Bond(in millions)Percent of TotalVolume

1988 $119,367.8 $29,365.1 24.6%

1989 $125,530.5 $27,650.2 22.0%

1990 $128,045.8 $31,426.0 24.5%

1991 $173,071.5 $27,809.6 16.1%

1992 $235,413.1 $26,868.0 11.4%

1993 $293,052.2 $21,230.6 7.2%

1994 $165,101.3 $25,054.9 15.2%

1995 $161,817.1 $27,942.7 17.3%

1996 $185,207.4 $31,262.0 16.9%

1997 $220,671.7 $37,550.0 17.0%

1998 $286,816.9 $46,288.0 16.1%

1999 $227,740.5 $47,408.0 20.8%

2000 $200,880.0 $41,392.0 20.6%

2001 $288,082.9 $49,356.0 17.1%

2002 $358,568.6 $50,248.0 14.0%

2003 $383,559.3 $45,633.0 11.9%

2004 $359,717.2 $47,877.0 13.3%

2005 $408,260.0 $54,691.0 13.4%

Sources: The 1988 to 1995 data are from Nutter, Sarah,Tax-Exempt Private Activity Bonds, 1988-
1995,” SOI Bulletin, Summer 1999; Belmonte, Cynthia, “Tax-Exempt Bonds, 2003-2004,: SOI
Bulletin, Fall 2006 ; and Belmonte, Cynthia, “Tax-Exempt Bonds, 2005, SOI Bulletin, Fall 2007.
Total long-term bond volume data are from the Bond Buyer Yearbook 2007.

6 Exempt facilities subject to the volume cap are the following: mass commuting facilities,
water furnishing, sewage treatment, solid waste disposal, residential rental projects, electric
energy or gas furnishing, local district heating or cooling provision, and hazardous waste
disposal and 25% of high-speed rail facility bonds. 26 I.R.C. Section 141(e), Section

142(a), and 146(g).

Bonds issued for several activities classified as private are not subject to the
volume cap if the facilities are governmentally owned.7 These activities are airports,
docks, and wharves; nonprofit organization facilities; high-speed inter-urban rail
facilities; and solid waste disposal facilities that produce electric energy. Table 7
below reports the estimated tax expenditure for selected private activities that qualify
for financing with tax-exempt debt.
Table 7. Federal Tax Expenditure for Selected Private Activities
Financed with Tax-Exempt Bonds
2007 TaxPercentage of
Private ActivityExpenditureTotal
(in millions)
Total of Selected Activities$8,310100.00%
Energy Facilities$300.36%
Water, Sewerage, and Hazardous Waste Disposal $3704.45%
Small-Issues $350 4.21%
Owner-Occupied Mortgage Subsidy$90010.83%
Rental Housing$8309.99%
Airports, Docks, and Similar Facilities$85010.23%
Student Loans$4405.29%
Private Nonprofit Educational Facilities$1,75021.06%
Hospital Construction$2,76033.21%
Veterans’ Housing$300.36%
Source: Office of Management and Budget. Analytical Perspectives: Budget of the United States
Government, Fiscal Year 2009, Table 19-1, pp. 287-291.
Recently, Congress has further expanded the types of private activities eligible
for tax-exempt financing and has increased the capacity for selected activities and
issuers. A brief description of legislation that Congress has enacted since 2001
follows below.
Economic Growth and Tax Relief Reconciliation Act of 2001. As part
of the Economic Growth and Tax Relief Reconciliation Act of 2001 (P.L. 107-16),
a new type of tax-exempt private-activity bond was created beginning on January 1,8

2002. The act expanded the definition of “an exempt facility bond” to include bonds

7 This does not mean governmental ownership in the conventional sense. It simply means
that lease arrangements for private management and operation of bond-financed facilities
must be structured to deny accelerated depreciation benefits to the private operator, lease
length must conform to the facility’s expected service life, and any sale of the facility to the
private operator must be made at fair market value. 26 I.R.C. Section 146(g).
8 For a more extensive explanation of the tax exempt bond provisions in EGTRRA, see CRS

issued for qualified public educational facilities. Bonds issued for qualified
educational facilities are not counted against a state’s private-activity volume cap.
However, the qualified public educational facility bonds have their own volume
capacity limit equal to the greater of $10 multiplied by the State population or $5
million. Since nearly all states would be better off with the $10 per capita limit, the
potential new debt would have been approximately $2.8 billion in 2001 if the bonds
were available in 2001.
Job Creation and Worker Assistance Act of 2002. The JCWA 2002
(P.L. 107-147) created the New York Liberty Zone (NYLZ) in the wake of the
September 11, 2001, terrorist attacks. The legislation included several tax benefits
for the NYLZ intended to foster economic revitalization within the NYLZ.
Specifically, the so-called “Liberty Bond” program allows New York State (in
conjunction and coordination with New York City) to issue up to $8 billion of tax-
exempt private-activity bonds for qualified facilities in the NYLZ. Qualified
facilities follow the exempt facility rules within section 142 of the IRC. The original
deadline to issue the bonds was January 1, 2005, but was extended to January 1,

2010, by P.L. 108-311.

American Jobs Creation Act. In 2004, the American Jobs Creation Act
(P.L. 108-357) created bonds for “qualified green building and sustainable design
projects.” The bonds are exempt from the state volume cap and are instead limited
to an aggregate of $2 billion for bonds issued between January 1, 2005 and October

1, 2009.

The Safe, Accountable, Flexible, Efficient, Transportation Equity
Act of 2005. This legislation created a new type of tax-exempt private activity
bond for the construction of rail to highway (or highway to rail) transfer facilities.
The national limit is $15 billion and the bonds are not subject to state volume caps
for private activity bonds. The Secretary of Transportation allocates the bond
authority on a project-by-project basis.
Gulf Opportunity Zone Act of 2005. The hurricanes that struck the Gulf
region in late summer 2005, prompted Congress to create a tax-advantaged economic
development zone intended to encourage investment and rebuilding in the Gulf
region. The Gulf Opportunity Zone, (GOZ), is comprised of the counties where the
Federal Emergency Management Agency (FEMA) declared the inhabitants to be
eligible for individual and public assistance. Based on proportion of state personal
income, the Katrina-affected portion of the GOZ represents approximately 73%of
Louisiana’s economy, 69% of Mississippi’s, and 18% of Alabama’s.9
Specifically, the “Gulf Opportunity Zone Act of 2005” (GOZA 2005, P.L. 109-

135) contains two provisions that would expand the amount of private-activity bonds

8 (...continued)
Report RS20932 Tax-Exempt Bond Provisions in the “Economic Growth and Tax Relief
Reconciliation Act of 2001”, by Steven Maguire.
9 See CRS Report RL33154, The Impact of Hurricane Katrina on the State Budgets of
Alabama, Louisiana, and Mississippi, by Steven Maguire.

outstanding and language to relax the eligibility rules for mortgage revenue bonds.
The most significant is the provision to increase the volume cap (see Table 3) for
private-activity bonds issued for Hurricane Katrina recovery in Alabama, Louisiana,
and Mississippi (identified as the Gulf Opportunity Zone, or “GO Zone”). GOZA
2005 would add $2,500 per person in the federally declared Katrina disaster areas in
which the residents qualify for individual and public assistance.
The increased volume capacity would add approximately $2.2 billion for
Alabama, $7.8 billion for Louisiana, and $4.8 billion for Mississippi in aggregate
over the next five years. The legislation defines “qualified project costs” that can be
financed with the bond proceeds as (1) the cost of any qualified residential rental
project (26 sec. 142(d)); and (2) the cost of acquisition, construction, reconstruction,
and renovation of — (i) nonresidential real property (including fixed improvements
associated with such property) and (ii) public utility property (26 sec. 168(i)(10)), in
the GOZ. The additional capacity would have to be issued before January 1, 2011.
The provision is estimated to cost $1.556 billion over the 2006-2015 budget10
The second provision allows for advance refunding of certain tax-exempt bonds.
Under GOZA 2005, governmental bonds issued by Alabama, Louisiana, and
Mississippi may be advance refunded an additional time and exempt facility private-
activity bonds for airports, docks, and wharves once. Private-activity bonds are
otherwise not eligible for advance refunding (see earlier discussion of advance
What Are Arbitrage Bonds?
Many individuals and businesses make money by engaging in arbitrage,
borrowing money at one interest rate and lending that borrowed money to others at
a higher interest rate. The difference between the rate at which one borrows and the
rate at which one lends produces arbitrage earnings. At its most basic level, it is the
primary activity of commercial banks — pay depositors an interest rate of “x” and
use the deposits to make commercial, automobile, and home loans at “x + y” interest
rate. In this context, arbitrage is a time-honored and appropriate financial market
That is not the case in the tax-exempt bond market. State and local governments
do not pay federal income tax, and absent federal constraint, have unlimited capacity
to issue debt at low interest rates and reinvest the bond proceeds in higher-yielding
taxable debt instruments, thereby earning arbitrage profits. Unchecked, state and
local governments could substitute arbitrage earnings for a substantial portion of their
own citizens’ tax effort.

10 The 10-year revenue loss estimates for GOZA 2005 are from the Joint Committee on
Taxation, Estimated Revenue Effects of H.R. 4440, the’Gulf Opportunity Tax Relief Act of

2005,’ as passed by the House of Representatives and the Senate on Dec. 16, 2005, JCX-89-

05, Dec. 20, 2005.

Congress has decided that such arbitrage should be limited, and that tax-exempt
bond proceeds must be used as quickly as possible to pay contractors for the
construction of the capital facilities for which the bonds were issued. Since it is
impossible for bonds to be issued precisely when contractors must be paid for their
expenses incurred in building public capital facilities, the tax law provides a
three-year period to spend an increasing share of the bond proceeds. Bond issues that
have unspent proceeds in excess of the allowed amounts during this three-year spend-
down schedule must rebate any arbitrage earnings to the Department of the Treasury.
Bond issues are considered to be taxable arbitrage bonds if a governmental unit, in
violation of the arbitrage restriction in the tax code, invested a substantial portion of
the proceeds “to acquire higher yielding investments, or to replace funds which were
used directly or indirectly to acquire higher yielding investments.”11
What Are Tax Credit Bonds?
The 1997 Taxpayer Relief Act created a new category of tax preferred state and
local bonds, the qualified zone academy bond (QZAB) for renovating public school
facilities. Congress authorized QZAB debt of $400 million a year for 1998 through
2005 or $3.2 billion over the eight years. The annual limit is allocated among the
states in proportion to their share of all individuals below the poverty line.12
In 2005, Congress created two new types of tax credit bonds called Clean
Renewable Energy Bonds (CREBs) and Gulf Tax Credit Bonds (GTCBs). CREBs
were created with the enactment of the Energy Policy Act of 2005, P.L. 109-58.13
The bonds are intended to encourage the development of renewable energy sources.
As with QZABs, purchasers of CREBs receive federal tax credits in lieu of interest
payments, and the borrower only repays the principal. The total CREB bond volume
is limited to $800 million nationally, and the bonds must be issued before January 1,
2008. Only $500 million may be allocated to qualified projects of qualified
borrowers “...which are governmental bodies.”14 In addition, a third party must spend
at least 10% of the proceeds of the issue for the bond to qualify. The third party
would likely be a private business or individual that would benefit from the CREB
After Hurricane Katrina, Congress created yet another tax credit bond called
Gulf Tax Credit Bonds (GTCBs). These tax credit bonds are different from the
others in that the maturity is capped at two years, and only three states can issue a
limited amount: Alabama ($50 million), Louisiana ($200 million), and Mississippi
($100 million). The tax credit rate will approximate the interest rate on equivalent
two-year taxable securities. In addition, GTCBs require an equal match from the
issuing state. The proceeds from the bonds can be used to pay principal, interest, or

11 26 I.R.C. Section 148(a).
12 For a more detailed explanation of tax credit bonds or QZABs, see CRS Report RS20606,
Tax Credit Bonds: A Brief Explanation, by Steven Maguire.
13 119 Stat. 594.
14 119 Stat. 994.

premiums on state governmental bonds or make loans to any political subdivision of
the state to then be used to pay principal, interest, or premiums on the political
subdivision’s governmental bonds. The bonds can only be issued in 2006.
The subsidy for tax credit bonds is not provided in the form of the exemption
of interest income from federal income tax, as is true for tax-exempt bonds, including
bonds issued for public elementary and secondary education facilities. Rather, the
subsidy for them is a credit taken by eligible financial institutions against the federal
taxes they owe. The credit rate is calculated by the Treasury Department such that
the bonds can be issued without discount and without any interest cost to the issuer.
Unlike QZABs, where issuers are chosen by the state, issuers of CREBs are selected
by the Secretary of Treasury. For GTCBs, the governor of the issuing state
designates the use of the bond proceeds.
One way to think of this alternative subsidy is that financial institutions can be
induced to purchase these bonds if they receive the same after-tax return from the tax
credit bonds that they would from the tax exemption. The value of the credit must
be included in taxable income, but is then used to reduce regular or alternative
minimum tax liability. Assuming the taxpayer is subject to the regular corporate
income tax, the credit rate should equal the ratio of the purchaser’s forgone market
interest rate on tax-exempt bonds divided by one minus the corporate tax rate. For
example, if the tax-exempt interest rate is 6% and the corporate tax rate is 35%, the
credit rate would be equal to 0.06/(1-0.35), or about 9.2%. Thus, a financial
institution purchasing a $1,000 tax credit bond would receive a $92 tax credit for
each year it holds the bond.
The implicit subsidy is much greater for tax credit bonds than for tax-exempt
bonds. All of the interest costs for tax credit bonds are paid by the federal taxpayer.
For tax-exempt bonds, the federal taxpayer absorbs only the difference between the
taxable and tax-exempt interest rates. For example, if the taxable rate is 9.2% and
the tax-exempt rate is 6%, the non-tax credit bond receives a subsidy equal to 3.2
percentage points, the difference between 9.2% and 6%. The tax credit bond receives
a subsidy equal to all 9.2 percentage points.
The relationship between three types of bonds, all with the same after-tax yield
of 6%, is presented in Table 8. The most important column is the last, titled
“Implicit Subsidy to Borrower,” which is synonymous with the federal revenue loss
presented in Table 2 for all outstanding tax-exempt debt. The implicit subsidy is the
amount the borrower saves because of favorable federal tax treatment. In the case
of tax credit bonds, the subsidy is considerably greater than with traditional tax-
exempt bonds.

Table 8. Comparison of Three Types of Bonds
with a 6% After-Tax Yield
(numbers are in percentage of bond proceeds assuming a corporate
taxpayer is in the 35% marginal tax bracket)
Type ofBorrowerLenderTaxes PaidFederal TaxRevenueImplicitSubsidy to
BondPaysReceivesby Lenderfrom BondBorrower
T a xable 9.2 9.2 3.2 +3.2 0.0
Tax Credit0.09.2a3.2b-6.0b9.2
a. The lender is allowed to reduce its tax liability by the amount of the credit.
b. The federal government receives some additional tax revenue because the credit is included in
taxable income. However, the loss in revenue from the credit exceeds the revenue gain
producing the negative revenue effect.
Legislative Issues
Current legislative interest focuses on altering the tax treatment of state and
local debt to provide even greater financial support for a variety of public projects
such as education infrastructure, healthcare facilities, and rapid transit. There are
several ways the federal government can increase the federal subsidy to state and
local governments that issue tax-exempt bonds for these targeted purposes. One,
arbitrage rules could be loosened to allow the state or local government to earn more
investment income on unused bond proceeds. Less stringent arbitrage rules may also
reduce the compliance burden of some smaller government entities freeing more
funds for the project. Two, the federal government can change the rules for private
activities that qualify for tax-exempt financing. If more activities qualified to use
tax-exempt debt finance by virtue of any proposed rules changes, previous
congressional efforts to limit the use of tax-exempt debt for non-governmental
projects would be mitigated. Third, Congress could introduce a new type of tax
favored instrument, such as tax credit bonds, for a broader range of activities.
In addition to the three options highlighted above, many other methods can be
employed to enhance the federal subsidy for state and local government capital
formation. The desire to subsidize state and local capital formation, which in many
cases may be justified, must be weighed against the federal revenue loss and the
potential for misallocation of federal tax revenue.

Suggested Readings
Ballard, Frederic L., Jr. ABCs of Arbitrage: Tax Rules for Investment of Bond
Proceeds by Municipalities (Chicago: Urban, State, and Local Government Law
Section, American Bar Association, 1992).
Describes and explains arbitrage and the arbitrage tax rules that control state and
local investment practices.
Congressional Budget Office. The Tax-Exempt Financing of Student Loans
(Washington: GPO, August 1986).
Provides a legislative history of student loan bonds, describes the operation of
student loan authorities, estimates the costs of student loan bonds, and discusses
policy alternatives.
Congressional Budget Office. Small Issue Industrial Development Bonds
(Washington: GPO, September 1981).
Describes the growth, uses, and effects of small-issue IDBs, and discusses
policy alternatives.
U.S. Government Accountability Office. Home Ownership: Mortgage Bonds Are
Costly and Provide Little Assistance to Those in Need. GAO Report RCED-88-

111, March 1988.

Discusses the structure and operation of mortgage revenue bonds and analyzes
whether the bonds are successful in increasing home ownership for the target
population of lower-income households.
Hilhouse, Albert M. Municipal Bonds: A Century of Experience (New York:
Prentice-Hall, 1936).
The classic history of the use and development of municipal bonds from their
introduction in the 19th century.
Petersen, John E., and Ronald Forbes. Innovative Capital Financing (Chicago:
American Planning Association, 1985).
Provides discussion of the numerous variations on tax-exempt bonds that have
been developed to raise capital for the state and local sector, such as sale-
leasebacks, installment purchase contracts, etc.
Public Securities Association. Fundamentals of Municipal Bonds (New York:
Public Securities Association, 1987).
Describes the structure and functioning of the state and local debt markets.

U.S. Senate. Committee on the Budget. Tax Expenditures: Compendium of
Background Material on Individual Provisions. S.Prt. 106-65. 106th Congress,

2nd session, December 2000.

Provides description, revenue loss estimate, and economic analysis of the effects
of governmental bonds and each major category of private-activity bond.
Zimmerman, Dennis. The Private Use of Tax-Exempt Bonds: Controlling Public
Subsidy of Private Activity (Washington: The Urban Institute Press, 1991).
Provides institutional background: history, legal framework, and industry
characteristics. Provides discussion of tax-exempt bonds as an economic policy
tool affecting: intergovernmental fiscal relations, the federal budget deficit,
efficient resource allocation, and tax equity. Provides a history and economic
analysis of tax-exempt bond legislation from 1968 to 1989.