Tax expenditures are government revenue losses resulting from provisions in the tax code that allow a taxpayer or business to reduce his or her tax burden by taking certain deductions, exemptions, or credits. Tax expenditures have the same effect on the federal budget as government spending. They can have effects on recipients similar to grants or other types of subsidies. For instance, if the government wants to encourage people to buy solar panels for their homes, it can either send checks to those who promise to buy the panels or offer tax breaks once the panels have been purchased.

Tax expenditures can affect more than just the targeted activity. When certain people or organizations are selected to receive targeted tax breaks through tax subsidies, the size of the tax base is reduced and tax rates then have to be increased for everyone in order to bring in an equivalent amount of revenue to the pre-tax expenditure level. Further, if a tax subsidy is not expressly intended to make a tax more efficient, then it will most likely produce an economic inefficiency. For example, when a tax subsidy is given to businesses to invest in a specific commodity, private investment is shifted from some other commodity into the tax-preferred area of investment without regard to the return on investment. This creates an economic inefficiency. Tax subsidies can also end up rewarding taxpayers for behavior they would have engaged in regardless of the tax benefit.

Because the federal income tax is a progressive tax and those with higher incomes pay a higher proportion in taxes, the value of a tax subsidy grows as income rises, reducing progressivity. This has led some to call tax expenditures "upside down subsidies," since they tend to generally benefit those with higher incomes more than those with lower incomes in a progressive income tax system.

Source: Analytical Perspectives, President’s Fiscal Year 2010 Budget.

Not only do tax subsidies benefit a small group of interests — often those with higher incomes — they usually do not go through open, transparent political processes. Particularly true in the case of tax subsidies, expenditures through the revenue code are often large, hidden, and not subject to the same public debate as direct spending. Further, they do not get evaluated like much direct government spending does. Politicians often employ tax expenditures because they can use the tax code to confer benefits to constituents while campaigning on a platform of lowering taxes, which sounds quite a bit different than explaining that they use the tax code to benefit certain people and not others.

There are experts who disagree on what baseline should be used to measure tax expenditures. Some assume that the current tax system is, or should be, a broad-based income tax. Others argue it is, or should be, a consumption tax. The baseline one uses can greatly change what gets counted. For instance, using an income tax baseline, the exemption for investment income from Individual Retirement Accounts would be considered a tax expenditure. Using the consumption tax baseline, this provision would not be counted as a tax expenditure since income from investments is not subject to a consumption tax.

Subsidyscope does not take a position on which tax structure is appropriate but presents tax expenditures estimates based on the income tax baseline because the Department of the Treasury and the Joint Committee on Taxation (JCT) use this baseline in making their estimates.

However, there are still differences between the estimates produced by the JCT and the Treasury even though the same income tax baseline is used. As the JCT explains in a recent document entitled "A Reconsideration of Tax Expenditure Analysis," the Treasury's analysis has changed over time. It has shifted from the use of a "normal" to a "reference" baseline, as the tax starting point from which to measure income tax expenditures. A "normal" baseline is considered more theoretical in nature, and based on the economic notion of a comprehensive income tax that applies to all income, which we have never had in practice as some types of income are excluded from taxation. Alternatively, a "reference" baseline more closely matches the actual tax code that we have, and considers some of the deviations from a normal income tax as part of the baseline rather than counting them as tax expenditures.¹

Illustrating the dynamic nature of this debate, the JCT proposed in 2008 to move away from the use of a "normal" income tax baseline toward the use of a "reference" baseline. This would incorporate current codified tax law and would more closely resemble Treasury's model, rather than comparing current law to some theoretical and heretofore undefined in practical use, "normal" baseline. JCT asserts that this revised approach will result in a more principled and neutral approach to the issues.

Based on this information, Subsidyscope will present the Treasury's estimates, which are compiled annually by the Office of Management and Budget in the Analytical Perspectives of the President's Budget. Those interested in the JCT estimates may find them here; for the transportation sector, the estimates are very similar.

Tax Expenditures for Transportation

Tax expenditures are not heavily used in the transportation sector in comparison to direct expenditures. As the table below illustrates, the totals in a given year are around $4 billion. Just as there are difficulties measuring subsidies in general, estimating tax subsidies is no different. Tax expenditure data presented are estimates of revenue forgone. They represent the lost revenue attributable to the use of the provision, which is not necessarily the same as what would be raised if the tax expenditure were repealed. Summing tax expenditures, while not technically accurate, often provides a reasonably good estimate for the total cost of groups of tax expenditures. The repeal of any single tax expenditure can trigger behavioral effects that in turn affect other tax expenditure amounts or even the total amount of tax revenue flowing into the Treasury. For example, if the tax expenditure favoring employee parking is repealed, more taxpayers may take the tax expenditure for employee transit passes, thus increasing the estimate for that tax expenditure.

1. See pages 25-27 of this document for a summary of the methodological variation between the Congress' Joint Committee on Taxation and the estimates produced for the President's Budget by the Treasury Department.