In proposing to renew the Build America Bonds program, the Obama administration last week set an audacious goal: more benefit at the same cost.
As explained in numerous papers and reports the past few years, the federal government believes the tax exemption on state and local government debt is a costly subsidy that siphons to investors much of the largesse intended for municipalities.
By shifting more municipal borrowing into the taxable market and tinkering with the subsidy rate, the federal government hopes, as it stated in its budget proposal, to “provide more support to state and local governments than do tax-exempt bonds, but at the same cost to the federal government.”
This statement is not as simple as it sounds. Refashioning the cost and support of the tax exemption in the guise of cash entails guesswork because both the cost of and the support provided by tax-exempt bonds are anything but clear.
A total of roughly $530 billion a year is spent on infrastructure projects like highways, airports, water and energy utilities, garbage disposal sites, schools, and hospitals, according to a joint study conducted last October by the Congressional Budget Office and Joint Committee on Taxation.
A bit less than half of this is spent by the private sector, and about an eighth is spent by the federal government itself, either through direct expenditures or through grants.
The remaining $215 billion in infrastructure costs are borne by state and local governments.
Since 1913 — which not coincidentally was when the income tax was enacted — the federal government has indirectly contributed to the municipal share of this spending by exempting interest income on state and local government debt from federal income tax.
This enables municipalities to borrow more cheaply, because investors are willing to accept lower yields on bonds when they are tax-free.
The benefits of this subsidy are the reduction in borrowing costs. The costs are the taxes the federal government does not collect. Washington believes the costs outweigh the benefits.
In its latest budget, the administration estimated it will sacrifice about $38.7 billion in tax revenue next year by excluding the interest on municipal bonds from taxation.
According to estimates cited by the CBO, only about 80% of this — or $31 billion — will actually benefit municipalities. The rest is waste.
In other words, reducing municipalities’ borrowing costs by 80 cents through tax exemption costs the federal government $1.
For example, say a school district wants to borrow $10 million for 30 years. Assume investors in the taxable bond market demand 6% a year from the district.
Under basic economic theory, the district’s borrowing costs in the tax-exempt market should be 3.9%, because investors in the 35% tax bracket would view a 6% taxable rate and a 3.9% tax-free rate as essentially the same thing.
If that were all there was to it, nothing would be wasted and this article would already be over.
In a reality divorced from the economic models, it does not happen that way. The reason: with municipalities floating more than $400 billion in bonds a year, there simply are not enough people in the top tax bracket to buy it all. Therefore state and local governments have to appeal to investors in lower tax brackets, too.
After exhausting the 35% tax bracket investors, the school district has to satisfy investors in the 25% tax bracket to find enough buyers for the deal.
That means it needs to offer at least 4.5% interest on the tax-exempt bonds, or else the 25% tax-bracket investors would prefer a 6% taxable rate to the one offered on tax-exempt bonds.
If the 4.5% rate is the one that clears the market, the district must pay that rate to everyone — even the top tax-bracket investors who in this example would have been willing to settle for 3.9%.
Now imagine the school district has borrowed $10 million at a 4.5% tax-exempt interest rate. The district has saved money compared with the 6% it would have paid in the taxable market. But the federal government has overpaid.
Theoretically, investors in the 35% tax bracket would have been willing to buy the bonds at a 3.9% rate — in fact that’s effectively what they would be collecting if they bought the 6% taxable bond. They instead are enjoying a 4.5% rate — free money.
Those extra 0.6 percentage points are dead weight: money the investors would have been willing to pay in taxes that instead is foregone and confers no benefit to the borrower.
Some of the subsidy intended for the municipality went to the bondholder.
Some analysts estimate the tax bracket of the “market-clearing” bondholder is sometimes as low as 13%. Every muni bondholder above that rate is enjoying federal largesse targeted for the municipality.
A Better Way?
For at least four decades, the federal government has tried to find alternatives to the tax-exemption subsidy, including a variety of tax-credit and subsidy proposals. None stuck until now.
The latest and probably most successful was delivered through the tax-credit and subsidized taxable debt authorized by the American Recovery and Reinvestment Act last year.
Municipalities have already sold more than $70 billion in taxable Build America Bonds, according to Thomson Reuters. Some market watchers forecast $130 billion to $150 billion in issuance this year.
That means a third of municipal borrowing could go through the taxable market, a significant shift for an industry that has been more than 90% tax-exempt almost every year. One might expect that paying the municipality directly instead of through the tax exemption would eliminate the inefficiency.
It would, except many people believe the 35% subsidy is too high.
Even though the subsidy may be more efficient, it results in the government paying more overall than the initial estimates with the legislation projected. Shortly after the measure passed, the Joint Committee on Taxation estimated the program would cost the federal government $292 million in fiscal 2010. The Treasury Department’s budget request was $340 million.
The numbers greatly underestimated the extent to which municipalities would utilize BABs. In its budget request for 2011, the Obama administration estimated the BAB program will cost the federal government $13 billion a year by 2020.
The latest budget proposal predicts a cost of $64 billion over the next decade.
The rush of state and local governments into the taxable market shows that they find the 35% subsidy more valuable to them than the tax exemption.
Throughout much of 2009, municipalities found that their after-subsidy borrowing costs through BABs were lower than borrowing costs in the tax-exempt market.
For instance, the New Jersey Turnpike Authority last April floated $1.7 billion in bonds, some of which were tax-exempt and some of which were taxable BABs.
The tax-exempt portion maturing in 2040 priced to yield 5.35%, while the taxable portion maturing the same year yielded 7.41%.
After a 35% subsidy, the taxable portion effectively cost the Turnpike 4.82%, representing more than half a percentage point of savings versus tax-exempt borrowing. While some of the discrepancy was attributable to a call option in the tax-exempt deal, the authority realized significant savings by utilizing the BAB subsidy and selling bonds in the taxable market.
If the federal government wants the revenue forgone through the tax exemption to equal BABs’ cash subsidy, it needs to estimate several things.
One is the average tax rate paid by tax-exempt bondholders. That would determine the lost tax revenue from not collecting taxes on tax-exempt debt. The CBO estimates that tax-exempt yields on average are determined by investors with a “market-clearing” tax rate of 21%.
The 35% cash subsidy to issuers is therefore of greater value than the tax exemption — while the tax exemption only reduces municipal borrowing costs by 21%, the subsidy reduces them by 35%.
But getting to the cost balance sought by the Obama administration is not as simple as resetting the subsidy rate to 21%.
Some of the subsidy is recovered from collecting taxes on BABs. Because the subsidy represents a portion of the interest on a taxable bond, the government collects part of the subsidy back.
While BABs have attracted investors that do not pay federal taxes, such as pension funds and other retirement accounts, some taxable bondholders do pay taxes.
That means the subsidy can be higher than 21% to match the cost of the tax exemption. How much higher depends on the tax rates paid by holders of BABs. With the strong possibility that federal income taxes will go up next year, the true cost of BABs to the federal government is a moving target.
The original goal of the program may not have been to replicate the value to the municipality of the tax exemption, but rather to exceed it.
When Congress and the administration were crafting ARRA, the federal economic stimulus package, municipalities had just suffered through a terrifying period. The bond insurance industry that had homogenized the product for investors had imploded. In the month after Lehman Brothers declared bankruptcy, municipal borrowing costs spiked a percentage point and a half, according to Municipal Market Data.
The volume of bonds floated by municipalities tumbled by a third in the fourth quarter of 2008, based on figures from Thomson Reuters.
Chris Mier, managing director at Loop Capital Markets, said the goal at first was clearly to deliberately overshoot the value of the tax exemption to liquefy the market for municipalities.
The market was in crisis and the government wanted to ensure the program’s use, he said, so it set the subsidy at a level that was comfortably high at the time.
Now that the crisis is over, Mier said there is less need for the BAB subsidy to so clearly outweigh the tax exemption.
“You could decrease that 35% and still induce people to issue BABs,” he said. “If you’re the federal government and you’re satisfied that this program is successful but you want to reduce the cost, you’d want to maybe fine-tune the BABs subsidy rate to save the federal government a little bit of money.”
The administration’s budget contends a subsidy at 28% would “be approximately revenue-neutral in comparison to the federal tax cost from traditional tax-exempt bonds.”
If the program was a salve for a crisis-stricken industry, and the crisis is over, then why the proposal to extend the program at all?
The budget claims BABs can replace an inefficient subsidy with an efficient one.
Mier, though, said some of the speculation is a little more nuanced than helping municipalities access the market or fine-tuning an inefficient subsidy into a less wasteful alternative.
Once municipalities are hooked on a cash subsidy for their borrowing in the taxable market, according to Mier, some people believe Washington will try to influence what projects are pursued by adjusting the subsidy level to favor certain types of projects.
“The first theory is, to save money,” Mier said. “The second theory is maybe the federal government wants to apply different subsidy rates for different types of bonds, depending on what they want to promote.”
Sen. Ron Wyden, D-Ore., who introduced BAB legislation in February 2005 and is considered a key supporter of the bonds, stressed that kind of refinement as talk of extending the program began late last year.
“I would like to see different flavors of BABs created,” he said. “That would allow us to adjust the subsidy and give, for example, transportation infrastructure investment a larger subsidy than other types of projects because transportation projects typically create more jobs and other public benefits.”
Will it Work?
The benefit of utilizing the 35% BAB subsidy compared with the tax exemption already has narrowed considerably.
Based on trades reported through the Municipal Securities Rulemaking Board, the BABs advantage the N.J. Turnpike would enjoy if it floated last year’s bonds today has shrunk to 24 basis points. Considering that the tax-exempt bonds embedded a call option — which are valuable to the issuer — and the BABs did not, that might not even be an advantage.
Cutting the subsidy would squeeze the discrepancy even more, in some cases erasing it.
George Friedlander, municipal strategist at Morgan Stanley Smith Barney, wrote in a report at the end of January that a subsidy of 28% would likely chase a significant amount of issuance back into the tax-exempt market.
The BAB subsidy would likely be more valuable to municipalities than the tax exemption only at much longer maturities, he said.
Some analysts have another concern. Allowing municipalities to access new sources of borrowing even as their tax receipts shrink, retiree health care costs skyrocket, and pension liabilities rise, might be a little like showing a fat person a skinny mirror just before he decides to go on a diet.
Richard Ciccarone, director of research at McDonnell Investment Management, worries that if the BAB program continues to ensure low-cost municipal borrowing, it could obscure investors’ signals about state and local government credit.
While he acknowledges the tax-exempt bond market has never done a perfect job of imposing discipline on municipalities with deteriorating credit, he said tax-exempt yields can still be meaningful.
Among other things, BABs introduce municipalities to new types of investors and foster a perception — justified or not — of an association with the federal government.
The resultant decline in municipal borrowing costs “may actually be a back-door approach to prolonging structural problems,” Ciccarone wrote in his commentary for 2010.
“BABs may open the door to shifting responsibility to Washington and indirectly contribute to helping many troubled state and local governments defer their tough decisions on unresolved structural fiscal issues that they might otherwise be forced to confront,” he said.