Conference Report Limits on Interest Deductions
December 17, 2017
The House and the Senate tax bills restricted the deduction of business interest. Each bill limited the deduction to 30 percent of “modified income” with carryover of the excess to later years. The Senate definition of “modified income” was much lower than in the House bill, and it more dramatically undercut investment incentives provided by the expensing provision. Unfortunately, the compromise in the conference bill takes the less damaging House version for four years, then switches to the more damaging Senate version for the future. This may have been done to raise revenue beyond the budget window to meet the Byrd rule, or to preserve other deductions. It was not done on tax policy merits.
The House definition is gross income, or EBITDA, which stands for earnings before interest, taxes, and depreciation and amortization. Firms may deduct up to 30 percent of EBITDA before facing the limit. The Senate bill puts the limit at gross income less depreciation and amortization, or EBIT. Firms must subtract depreciation and amortization to calculate the income ceiling.
- This is a lower number, and the interest deduction would be more constrained. (Thirty percent of a smaller number is a smaller number.)
- Worse, if a firm increases its investment, its depreciation rises, and the limit becomes tighter. The act of investment may cost a firm a portion of its interest deduction.
- Worse yet, this might happen even if the new investment is done with cash from current sales, or repatriated profits from abroad, or proceeds from new share issues, without any new borrowing. It would penalize past investment funded by old debt.
- Worst of all, it undercuts the investment incentive of the expensing provision for equipment. It also undermines the Senate bill’s granting of a lower, 25-year asset life to new buildings and other structures, which would do wonders for construction, factory creation, transportation, and urban redevelopment.
The Senate’s limitations on depreciation may stem from the erroneous concern, prevalent in tax circles, that one should not allow interest deductions when expensing is available, for fear of “negative tax rates.” This concern comes from a misunderstanding of the size and nature of returns on ordinary investment, and disregard of the taxes paid by the taxable lenders who set interest rates in the market. The earnings stream from the investment is taxed appropriately, some on the borrower’s tax return, some on the lender’s. Let us review the concern and why it is misguided.
There are two consistent ways to treat interest in an economically optimal tax system, one that does not discriminate against saving and investment. Either the interest should be deducted by the borrower and taxed to the lender, or the interest should not deductible by the borrower and not taxable income for the lender. In either case, a portion of the returns on an investment that is financed with borrowed money is subject to one layer of tax, paid by either the borrower or the lender but not both. Taxing the interest portion of the investment return on the borrower’s tax form by disallowing the deduction, and taxing it again when the lender receives it, is double taxation. This is akin to the double tax imposed on dividends or retained earnings of C corporations (where, in addition to the corporate tax, the shareholders are taxed when the dividends are paid, or taxed later as retained earnings raise share prices and the shareholders take the resulting capital gains). This double taxation of C corporations is one of the biases in the income tax against saving and investment.
Treating all sources of financing alike?
Removing the deduction for interest would eliminate the distinction between debt and equity finance for C corporations. However, it would introduce a new distortion between debt financing and financing with a business’s own income for the entire pass-through sector. That sector currently generates half of all business income. Elimination of the interest deduction without excluding interest from tax would be an enormous new distortion impeding capital formation. It would raise the cost of capital for the affected businesses, reduce returns to the owners, and discourage capital formation.
The fear that the current difference in taxation between debt and equity encourages “too much debt” is overblown. It applies only to the C corporation sector, and firms’ ability to add debt is limited. As a corporation becomes more highly leveraged (adds more debt), lenders regard the firm as a riskier borrower, and raise the interest rate they charge the firm. The business refrains from adding new debt when the higher charges on the debt service exceed the cost of obtaining funds by issuing new equity. The problem is self-limiting, and of little economic consequence.
Negative tax rate argument
Some critics of full expensing of the cost of capital equipment and buildings argue that it creates a subsidy for an investment financed with borrowed money. They argue: “If a business can expense its capital costs, the tax rate on the investment is zero. If the business can also deduct interest, it is paying a negative tax on that investment; this constitutes an investment subsidy from the government.”
The argument then goes: “All profits are eventually competed away, so businesses only earn returns on their assets that are barely equal in value to the cost of the assets. Therefore, if a firm gets to deduct the full cost of the assets on its tax return, it will have zero reportable taxable income over the life of the investments, and pay no tax. That is OK if it buys the asset with its own (after-tax) money. But if it borrows money to buy the asset, and gets to deduct the interest, it will have a negative tax on the projects due to the interest deduction. Therefore, if we let the business deduct the entire cost of buying the asset, we should deny the business a deduction for the interest it pays on the borrowed money.”
There are glaring errors with this analysis.
The initial premise is wrong. Not all profits are competed away. Assuming all profit is competed away is an unrealistic oversimplification popular among academics to facilitate their analysis. In the real world, assets normally earn more than they cost to be able to cover risk and to reward innovation. Expensing the cost of an asset does not wipe out taxable income, nor send taxes or tax rates to zero. There are several reasons why.
Producers with a particular edge will always earn outsized profits (called “quasi-rents”). A new and better product will yield unusually high returns until the competition can catch up. Also, no rational business would borrow to buy an asset if there were no return left over after paying the bank or the bondholder. Profits can be very low in some industries, but the need to cover the basic time value of money, plus some reward for the work of managing the enterprise, plus a significant cushion for risk, requires there to be some prospect of net income as the norm.
The claim that the deduction of the interest payment by the borrowing business (in addition to the cost of the asset) creates too much of a deduction and creates a negative tax is wrong. It ignores the tax paid by the lender. The lender receives some of the return on the asset in the form of interest and pays tax on that portion. The borrower pays tax on the rest.
But what if the lender does not owe tax? Some interest is collected in tax-deferred pensions and other types of retirement plans, and is not taxed immediately; but in these cases, the government collects the tax later, upon the distribution of the pensions, on an even larger amount of accumulated earnings as the reinvested savings build up and are paid out. Some small fraction of interest goes to tax-exempt charities and colleges, but the decision not to tax those entities was made by the government, presumably for some good social policy reason, and it is those entities that are being subsidized, not the borrowers. Also, these donation-dependent entities have limited flexibility to alter their saving; they can only save what they are given by their donors, and the charities are required to distribute most of it each year. They cannot suddenly increase their investments, and therefore are not the marginal lenders who can step up with additional money to fund the growth of the capital stock.
Growth is funded by taxable lenders or shareholders, and interest deducted by a borrowing business is taxable to those interest recipients. The interest rates in the credit markets are set by the marginal lenders, who are taxable, and who demand sufficient interest to cover the tax on the income. Businesses must pay the lenders a tax-inclusive rate. That some non-marginal lenders are tax-exempt is no excuse for not allowing a business to deduct its tax-inclusive interest costs when it must borrow at interest rates that are set in the taxable portion of the credit market.
For all these reasons, the argument that expensing produces negative tax rates on investment and constitutes a subsidy in the presence of interest deductions is flat wrong. There is no good reason to eliminate interest deductions to permit expensing. Expensing is a key element of any tax system which seeks to put all economic activity on a level playing field.
 “Laws, like sausages, cease to inspire respect in proportion as we know how they are made” is a sentiment more correctly attributed to the poet John Godfrey Saxe than more commonly attributed to Otto von Bismarck.
 In the second case, with no deduction by the borrower but no tax on the lender, the interest rate would drop to the tax-exempt level. The borrower would pay the same rate, after-tax, as in the deductible case, and would be held harmless. The government would pay less interest on newly tax-exempt government bonds. It would also collect more revenue, because more borrowers than lenders are currently taxable. The losers would be the tax-exempt institutions, that now piggyback on the higher taxable market interest rates.
 Assets generally earn at least a bare bones return, greater than zero, and often much more. Anything over the “bare bones” return is called “economic profit,” and it is that which economists often claim is competed away, eventually. But zero “economic profit” still leaves the bare bones return to be entered in taxable income. Expensing would offset much of the tax on that bare bones return on depreciable physical capital, but it would not protect returns to risk, land, intellectual property, or the many other sources of returns to a business.