J.D. Foster J.D. Foster
Former Senior Vice President, Economic Policy Division, and Former Chief Economist


September 25, 2017


Eliminating net interest expense for a wide swathe of business taxpayers is one of the major revenue sources contained in the House GOP’s “Better Way” blueprint1 for comprehensive tax reform. This policy would represent a radical departure from traditional tax policy and income tax theory. Not surprisingly, the net interest proposal has raised many questions and left affected businesses scrambling to gauge its effects in the context of the other major elements contained in the committee blueprint.

This article takes no position on the net interest deduction proposal itself. Instead, its purpose is to dispatch a long-standing tax policy myth that may be raised by some to justify the proposal, or any other that would limit interest deductibility. Dispelling this specific myth should help focus the coming debate on real concerns and real strengths, the latter of which center on the substantial revenue raised to fund other reforms to render a substantially more pro-growth federal tax code. The myth is the proposition that a tax system allowing for expensing creates a subsidy for debt finance if businesses are also allowed to deduct interest expense.

This myth has been around for decades and apparently is taught religiously in tax law and public finance courses nationwide. Because debt often carries a connotation bordering on immorality, or at least financial imprudence, the lectures apparently often fall on receptive ears. As argued below, however, it remains just a myth whose time to fade away has finally come.

The issue of interest deductibility arises because a fundamental feature of all modern tax reform proposals is the move toward expensing from some system of accelerated depreciation of capital purchases or capital consumption allowances. Indeed, if the myth were anything but a myth, the tax code would already demonstrate a strong subsidy, as any form of accelerated depreciation relative to economic depreciation would then generate some element of subsidy. In reality, expensing and interest deductibility go together as neatly as peas and carrots.

Neutrality and the Matter of Interest

Tax reform can significantly improve economic performance by reducing or eliminating various tax distortions to economic decisionmaking. The simple intuition is that the economy generally performs best if individuals and businesses can make economic decisions without having to consider tax consequences. The pursuit of such a tax system is the pursuit of a “neutral” tax system.

Slightly more technically, a neutral tax system is one that leaves relative prices unchanged. Relative prices guide decision-making, if the relative price is whether to buy one product versus a competing product, to consume today or to save and invest, to work more or enjoy more leisure, or to invest in one location or another. The baseline for comparison is then an economic system of relative prices without taxation, and a more neutral tax is one that more perfectly preserves relative prices despite the imposition of tax. This definition of neutrality is presumably intuitive to most people, but it is rarely explicitly stated - an oversight that occasionally leads to confusion.

Interest income is generally subject to tax under current law as well as under just about every modern federal tax reform proposal. Taxation thus creates a “tax wedge” between the pretax and after-tax returns to saving and lending. Because the after-tax return is determined foundationally by global forces reflecting a general time rate of preference, the effect of taxing interest income is to raise the interest rate charged to borrowers. For example, assume an after-tax rate of interest of 3 percent and a tax rate on interest income of 25 percent. The market interest rate would then be 4 percent, the tax having created a single percentage point tax wedge.

The tax wedge is not merely a theoretical proposition. It can be observed in the marketplace every day in comparing the interest rates charged on tax-exempt debt, such as municipal bonds, with the rates charged for taxable instruments of similar maturity and risk.

Consider now the transaction from the borrower’s perspective. Absent taxation, the borrower would face a 3 percent rate. With the tax on interest income, the borrower’s interest rate is now 4 percent — the tax has created a distortion and a bias against borrowing. Now suppose the borrower’s tax rate is 25 percent and the borrower is allowed to deduct interest expense. The net effect is then to set the borrower’s after-tax borrowing cost at 3 percent — exactly the rate that would be charged absent a tax on interest income. The moral of the story is that the deduction for interest expense preserves tax neutrality in the presence of a tax on interest income.

Neutrality and Expensing

In theory, a pure Haig-Simons income tax relies on economic depreciation to guide the structure of capital consumption allowances. Over any period, productive capital tends to decline in value because of use or the simple passage of time. The measured decline in value is the “economic depreciation” that occurred in the period. An accrual-centric measure of economic income then charges the amount of economic depreciation against total receipts in each tax period. This income measure is perfectly appropriate for some purposes, but not for preserving tax neutrality. The effect of following economic depreciation is to include in the business income tax base the real, after-tax, riskless return to capital — a concept analogous to the after-tax return to saving described above.

The cost of capital is the pretax return required for a business to be willing to undertake a particular investment. The cost of capital reflects the effects of all relevant aspects of the tax system as they affect the returns to the business’s owners, including depreciation, other deductible expenses, the tax rate, special taxes levied on the business, and any taxes — such as a tax on dividends — levied on the owners directly resulting from their ownership position. The cost of capital is then analogous to the pretax interest rate described above.

Using economic depreciation as a guide to tax policy raises the cost of capital on investment, thereby creating a specific tax wedge on investment returns. As businesses over time face various investment opportunities, typically presenting a range of expected returns, and because the cost of capital is pushed up by tax policy, the number of economically viable investment projects declines. In this way, a higher cost of capital reduces the amount of capital used in the economy, thus reducing output, productivity, profits, and wages. A central goal of pro-growth comprehensive tax reform is to minimize the extent to which the tax system raises the cost of capital and thus deters capital formation.

The federal income tax has long used some form of “accelerated” depreciation. Accelerated depreciation means the business can still only deduct 100 percent of a capital asset’s purchase price, but it can deduct more of the price sooner, thus raising the discounted present value of the deductions taken.

Expensing, or allowing a 100 percent deduction in the year an asset is purchased and installed, has gained increasing currency since roughly the mid-1990s. Expensing acknowledges that there is no economic policy basis for taxing the real, riskless return to investment, and the adoption of expensing is one of the most effective ways of reducing the cost of capital and spurring capital formation.

With the foregoing as preparation, it is straightforward to observe why the combination of expensing and interest deductibility preserves tax neutrality. Interest deductibility preserves neutrality regarding debt finance as long as interest income is taxed, and this is true regardless of the capital consumption system used. Expensing preserves neutrality regarding the real, riskless return on investment. Combining interest deductibility and expensing thus likewise preserves neutrality as a matter of what physicists might call a “conservation” of neutrality.

A Simple Model of Interest and Expensing

Those who shudder at mathematical representations may prefer to skip to the conclusion; those who prefer their theory leavened by arithmetic discipline may take comfort in the conclusions presented with a few simple manipulations. For what follows, the expression F(K) denotes a simple profit function based on a production function demonstrating the usual properties when output varies with the amount of capital (K) used.

A straightforward representation of the business decision regarding how much capital to use involves the simple statement that the initial investment must, after some period, produce at least a minimal return reflecting the required after-tax rate of return plus an amount necessary to preserve the original principal. Framed in those terms, a formula representing the decision process can be shown simply by showing all the flows involved, properly discounted and properly taxed.

Consider the simplest case of a tax-free, oneperiod, equity-financed investment required to earn an after-tax rate of return r, and such that the capital used experiences economic depreciation over the period d. That type of decision process can be shown as follows, with the initial investment on the left-hand side and the discounted return on the right-hand side:

(1) K = F(K) + K(1 - d) 1 + r

The behavior of a profit-maximizing company can then be shown by differentiating both sides regarding K and rearranging terms to yield:

(2) F' (K) = r + d

Suppose now that the company is subject to tax at rate t and is allowed to expense capital purchases and that lenders are subject to tax on interest income, so that the discount rate is now R = r/(1 - t). The company’s investment decision is then described by:

(3) K - tK = F(K)(1 - t) + K(1 - d) - tK(1 - d)
1 + R(1 - t)

The term -tK on the left-hand side represents the tax savings from expensing, while the expression tK(1 - d) on the right represents the tax levied on the remaining value of the capital used as the investment is liquidated at the end of the period. The latter term is often neglected in such algebraic representations and is often the source of error. Differentiating (3) as before and rearranging terms leaves:

(4) F' (K) = r + d

The expression in (4) matches the result in (2), demonstrating that expensing preserves neutrality in the face of income tax.

Finally, consider three further and interrelated new elements. First, suppose the investment is debt financed, represented by D; that interest expense is deductible; and that only a fraction, a, of the value of the asset sold at the end of the first period is subject to tax. In this case, the expression in (3) becomes:

(5) K – tK – D = F(K)(1 – t) + K(1 – d) - taK(1 – d) - RD + tRD - D 1 + R(1 – t)

As the amount of debt issued, D, is equal to the amount of capital purchased, K, expression (5) can be simplified to:

(6) -tK = F(K)(1 – t) - Kd - taK(1 – d) - K(1 + R(1 – t)) 1 + R(1 – t)

Differentiating (6) as before and rearranging terms yields:

(7) F’ (K) = r + d - t(1 – d)(1 – a) 1 - t

Notice two aspects of the result in (7). First, the cost of capital in (7) when the company is subject to tax and is allowed expensing and a deduction for interest expense is lower than when no tax is imposed (2) by the amount of the last term relating to the possibility the asset has an untaxed residual basis exceeding its tax basis. This result indicates the issues raised when the model fails to reflect the fact that when an asset is sold, if its sale price exceeds the remaining depreciable basis, the difference must be included in taxable income, or else a bona fide tax subsidy results. However, the tax subsidy relates to the failure to tax remaining basis, not the interest deduction.

Second, even though the business can expense its capital purchase, the purchase is debt financed, and interest is deductible, as long as the sale of capital at the end of the first period is fully taxed (a = 1), the cost of capital is invariant to the tax system, which is the classic definition of tax neutrality. That is, under these conditions, the expression in (7) reduces to copy the expression in (2) of a no-tax world.

We have noted above one common source of error — the omission of the tax on the recapture of excess basis — often misleading to the conclusion that expensing and net interest cannot combine to yield a neutral tax system. A second common source of error misleading to an incorrect conclusion is a failure to modify the discount rate to recognize the tax levied on interest income. That is, the pretax return, R, is included in the expressions (3) through (7) rather than the aftertax return, R(1 - t).


As pro-growth tax reform is once again on the agenda, ancient myths ought not be allowed to bias the final product. Whether one prefers referring to improving efficiency, increasing tax neutrality, or some other expression as the guide to pro-growth tax reform, if successful, the net result is generally the same: a substantial increase in tax neutrality and economic performance. This holds for the allocation of capital as well as the amount of productive capital used. Getting expensing right is a key part of improving neutrality for business investment. Whatever Congress decides regarding net interest, the motivation should not follow from myths about the relationship between interest expense, expensing, and neutrality.

About the authors

J.D. Foster

J.D. Foster

Dr. J.D. Foster is the former senior vice president, Economic Policy Division, and former chief economist at the U.S. Chamber of Commerce. He explores and explains developments in the U.S. and global economies.