Each year, the Joint Committee on Taxation (JCT) releases a report on the tax expenditures in the U.S. tax code. The report can be useful, particularly for looking at trends over time, but defining which tax provisions count as tax expenditures is a fraught process. As such, simply eliminating all tax expenditures would be misguided, but zeroing in on the report reveals where further scrutiny is warranted.

A tax expenditure is any part of the tax code that departs from “normal.” As a result, whether a tax provision is a tax expenditure depends on one’s definition of a “normal” tax code. JCT uses a Haig-Simons income tax as normal, defining income as consumption plus changes in net worth. This is problematic because a change in one’s net worth usually becomes consumption later, resulting in multiple layers of tax on some income. Alternatively, a saving-consumption-neutral tax solely focuses on consumption, resulting in one layer of tax.

Table 1 shows the largest corporate and individual tax expenditures in 2022 according to JCT.

On the individual side, the tax treatment of saving is the clearest illustration of the gap between the two systems. Under a pure income tax system, income that is invested is taxed twice: first when it’s earned and again when it produces a return. Thus, providing deductions for retirement savings account contributions, like 401(k)s, or exempting the returns, like Roth IRAs, is a tax expenditure under a pure income tax.

Under a saving-consumption-neutral tax system, each dollar is taxed only once, whether consumed immediately or saved for future consumption. Thus, retirement savings provisions would not be tax expenditures.

Given the limits on 401(k)s and other savings vehicles, our current tax code penalizes rather than subsidizes saving relative to consumption. This penalty exists because when you earn income and consume it immediately, you incur only one layer of tax, but if you choose to invest it to consume later, you incur two layers of tax.

A similar divide exists in business taxation, best illustrated in the context of deductions for investment costs. Consider a company that invests $100,000 in a new heavy truck with a useful life of five years. Under a pure income tax, the company’s net worth remains nearly unchanged, only declining by the economic depreciation (or wear and tear) of the heavy truck over the first year, and so the company is allowed a deduction for economic depreciation. Allowing companies to deduct investment costs faster than economic depreciation is thus considered a tax expenditure.

Conversely, under a saving-consumption-neutral tax system, the company would deduct the entire $100,000 investment in the new heavy truck immediately. Measuring corporate profits as revenues minus costs and including a full deduction for capital investment reflects the real costs incurred. By stopping short of full expensing, the tax code penalizes, rather than subsidizes, investment.

Current provisions for saving and investment are important structural elements to remove biases from the income tax system. Omitting provisions that ensure income is taxed only once, including defined-benefit and defined-contribution plans, traditional and Roth IRAs, reduced taxes on capital gains, and depreciation-related provisions, would change total expenditures and the composition of the largest tax expenditures significantly.

However, the removal of the saving and investment provisions above does not complete the task of isolating true tax expenditures.

For example, on the individual side, residential housing investment can face multiple layers of taxation, or zero layers of taxation, depending on the circumstance. Two provisions apply here: a capital gains tax exclusion for the sale of a personal residence, which is currently capped at $250,000 for single filers and $500,000 for joint filers, and the deduction for home mortgage interest on mortgages up to $750,000.

In practice, some residential housing can get the equivalent of both Roth and traditional 401(k) treatment. With the mortgage interest deduction, taxpayers can deduct the investment when they are making it (analogous to a 401[k]), and with the exclusion of capital gains, can exempt the returns from tax, analogous to a Roth IRA. Both provisions are capped, so only a portion of housing investment receives a double benefit.

International tax issues (namely the reduced tax rate on active income of controlled foreign corporations and the deduction for foreign-derived intangible income [FDII] derived from a trade or business within the United States) are also not clear-cut issues. The United States originally took a primarily worldwide approach to international tax issues but has moved partially towards a territorial system. As a result, certain provisions may qualify as expenditures relative to the original worldwide system, but they might not in the context of a territorial system. Additionally, the JCT estimates are the revenue cost of the provisions now, not the revenue raised from eliminating them. That is particularly worth remembering in the context of FDII, where the deduction is meant to encourage corporations to keep highly mobile intangible income in the United States.

When peeling back layers of the JCT report, it becomes clear that many tax expenditures are not “loopholes” or benefits for narrow special interests, but important structural elements of the tax code. After subtracting out provisions aimed at saving and investment, more than $1 trillion of expenditures remained in 2022 alone. Simply eliminating expenditures across the board would be misguided, but setting aside the important structural elements still leaves plenty of places to cut tax breaks.