Share |

Navigating The Tax Expenditures Minefield

e21 | 07/08/2011

Yesterday, there were reports that the chances of congressional leaders and President Obama reaching a long-term budget deal in the next few days were “maybe 50-50.” Members are now apparently considering three options: a small deal that would cut the federal deficit by $2 trillion over 10 years, a medium size deal that would trim $3 trillion, and then a “grand deal” that would shrink the deficit by $4 trillion. Until this week, Washington was focused on the contours of deal in the $2 trillion range, but was still at an impasse over tax policy. The relatively new prospects for a grand deal, however, will probably hinge on the elimination of “tax expenditures,” which are tax policy provisions that result in a reduction in income tax receipts for reasons unrelated to tax administration.

The Obama Administration has insisted that increases in revenue must be a part of a “debt-ceiling” compromise that cuts spending. Leaders in the House of Representatives and many Senators refuse to support any increase in revenues that result in new and higher marginal tax rates. Reducing tax expenditures, therefore, appears to be the only mechanism able to satisfy both sides, as it offers a way to generate additional revenue while leaving effective marginal tax rates unchanged.

The problem with this basic formula is that no one is sure what a tax expenditure really is. Everyone seems to agree that a “tax expenditure” is spending that takes place through the tax code through exclusions, deductions, and credits that are tantamount to government outlays. Unfortunately, this general definition does not square with the fact that the one area where actual spending through the tax code occurs – refundable tax credits, which allow households to receive checks from the government in situations where they owe no income tax – is not classified as a tax expenditure. The child tax credit, for example, costs about $370 billion over 10 years, but only $178 billion is a tax expenditure (p. 196). The other $192 billion is not foregone revenue in pursuit of policy, but instead direct outlays made by the IRS to households with no income tax liability (it gets counted as government spending). One might think that eliminating the direct payment to households that don’t pay taxes would be a way to cut “tax expenditures.”

In fact, such a policy would have no revenue impact and, as a result, would not be scored as reducing tax expenditures. To further complicate matters, the tax treatment of certain transactions is classified as a tax expenditure while the same treatment for other, closely related transactions is not. For example, the deductibility of interest on debt incurred as part of a business venture is not a tax expenditure, but the interest deducted on a similar loan secured against a second residence is. JCT explains the distinction (p. 39 or p. 43 of the pdf):

Under the general rules of the Code, individuals would be allowed to deduct only the interest on indebtedness incurred in connection with a trade or business or an investment. Thus, the deduction for mortgage interest on a principal or second residence is classified as a tax expenditure.

The confusion about tax expenditures in these cases stems from trying to separate what deductions (from income) are legitimate expenses or “general rules of the code” form those deductions that are special favors for certain groups or items that advance a certain policy agenda. Even the JCT and OMB disagree in this area: OMB classifies the imputed rental income from owner occupied housing as a tax expenditure (the implied rental income derived from living in your house), while JCT does not. JCT doesn’t contest that this is a tax expenditure, but argues that classifying it as such would make other housing-related tax expenditures, like the interest deduction, legitimate tax expenses.1 Conservatives generally contest the notion that the preferential rate on capital gains income is a tax expenditure because the savings invested to generate the capital gain has generally already been taxed once, suggesting the natural rate of taxation on capital gains should be zero. Similar disagreements abound.

Perhaps the diciest definitional issues arise when everyone agrees that a certain tax policy is a tax expenditure, but some lawmakers propose eliminating it in a way that is tantamount to a tax increase. This appears to be precisely the situation that exists today, as President Obama is proposing a reduction in the “rate” applied to deductions for certain higher income taxpayers. First off, it is important to realize that there is no “rate” applied to deductions, there is only a rate applied to income. As explained by e21 previously, the tax rate applied to the itemized deductions of middle-income families is exactly the same as the deductions made by high-income families – zero. The rate at which the deducted dollar would have been taxed is irrelevant for determining whether or not the deduction itself is legitimate. A deduction does not suddenly become an affront to economic-neutral tax policy because it is claimed by a household with a gross income in excess of $200,000.

More significantly, creating gross income thresholds for determining whether or not an item is a tax expenditure actually increases marginal tax rates. Perhaps the most insidious part about the tax code is the high marginal rates in income ranges where tax expenditures phase out. The chart below, calculated for the 2005 Tax Reform panel by the Treasury Department, shows that households with incomes below $30,000 face marginal tax rates above the top statutory rate of 35% because each additional dollar they earn reduces the earned income tax credit (EITC). Similarly, the marginal rate applied to the additional income generated by households with $80,000 of earnings increases because the child tax credit (in addition to other provisions, like the tuition credit and student loan interest deduction) begins to phase out. Rather than scale back tax expenditures, the Obama Administration’s proposal would have a similar effect: increasing marginal income tax rates on households at the arbitrarily defined income cut-off.

marginal rates

The odd thing about the Administration’s insistence on tax increases is that a simple elimination of some tax expenditures would generate more revenue without the impacting marginal rates and still fall mostly on upper-income taxpayers. The table below from the JCT looks at the distribution of benefits received under the mortgage interest deduction by household income. About 70% of the benefits are claimed by households with income above $100,000, while 8% of benefits go to households with $50,000 or less of income. According to OMB’s 2012 budget (see table further below), eliminating the mortgage interest deduction would raise $609 billion over five years. This would generate more than $1.2 trillion in revenue over ten years by removing a tax bias that would not raise anyone’s marginal income tax rate or impact decisions to work, save, or invest (with spending on housing viewed as a form of consumption).

The complexity of these definitional issues should in no way dissuade leaders in Congress from pursuing reductions in tax expenditures. This minefield can be navigated in a way that reduces deficits and the government’s footprint in the economy. Most significantly, the current revenue loss from existing tax expenditures is so great that negotiators need not even agree on every definitional issue. The table below, adapted from OMB’s 2012 budget, shows that the revenue loss from tax expenditures equals $1 trillion annually, or more than $6.4 trillion over 5 years.

Revenue Impact of Tax Expenditure

Conservatives are understandably reluctant to support any policy that results in less take-home income for households and businesses and more tax revenue for the government. But it’s important for policymakers to remember that in many cases tax expenditures are really no different than a direct government outlay for some household that is just financed by a different set of households.

A household that rents pays more in taxes because it cannot deduct its rental payment; a worker who buys her own health care policy pays more in taxes because she cannot deduct the cost of insurance acquired on the individual market; a business owner pays more in taxes when he sells his company because family businesses do not receive the special exemption from capital gains taxes afforded to principal residences. In each case, the disadvantaged household is effectively subsidizing the similarly situated taxpayer who happens to fall on the “right side” of the tax expenditure divide. Since many of these tax benefits are economically equivalent to direct subsidies, it is fair to view eliminating them as the same as cutting government spending. Let’s hope the confusion in this area does not ultimately preclude a grand deal in D.C. that includes substantial reductions in government outlays.


1. “If the imputed income from owner-occupied homes were included in adjusted gross income, it would be proper to include all mortgage interest deductions and related property tax deductions as part of the normal income tax structure, because interest and property tax deductions would be allowable as a cost of producing imputed income. It also would be appropriate to allow deductions for depreciation and maintenance expenses for owner-occupied homes.”