“Should 5 percent appear too small, be thankful I don’t take it all”

- “Taxman,” the Beatles

We appreciated the provocative, tongue-in-cheek panaceas that Michael O’Connor, in the September 2020 issue of this newsletter, proposed for the broader tax issues facing Canada—challenges that will only be more difficult to resolve following the pandemic, when the government needs more tax revenue. Imagine the sound of O’Connor’s soft tenor voice along Elgin Street, singing “Losing My [Interest] Deduction,” his simple love song to the abolition of interest deductibility for most Canadian businesses. Then imagine that he follows that up with his recent YouTube phenomenon, “If I Just Applied the GAAR,” and maybe throws in, for good measure, a musical cure for transfer pricing, with “My Heart Aches for Formulary Apportionment.” Tax returns will simplify, assessments will amplify, and collections will multiply. It surely will be the best of all possible worlds.

What could possibly go wrong? Denying the deduction of an interest expense may sound simple. But make no mistake: even if implemented with a general corporate rate decrease, it is nothing other than a tax on businesses that rely on borrowing. The Liberal and NDP election platforms recognized that their own proposals to limit corporate interest deductions to 30 percent of ebitda (earnings before interest, taxes, depreciation and amortization) would raise an estimated $12 to $21 billion in the next decade. (These proposals would exclude net interest expense less than $250,000, in order to reduce the impact of the change on small business.) O’Connor, by suggesting that his more extensive changes would protect the “tax base at a time when tax revenues are most needed by government,” implicitly recognizes that his changes will lead to tax increases, though he also suggests that the introduction of the interest denial could be accompanied by reductions in headline tax rates on a “revenue-neutral basis.”

Interest is an ordinary business expense

Where we fundamentally differ from O’Connor is in our view that interest is part of the determination of profit. He views interest as a distribution of profit, not as a cost of doing business. The owner of a business that borrows from a bank to buy an extra machine does not have more profit until the machine produces enough revenue to cover the costs associated with its acquisition and operation, including the cost of interest. The bank that makes the loan would hardly be considered to be sharing in the owner’s profit. For the owner of a business, loans rank ahead of distributions of profit, and when the loans are secured, they may rank ahead of ordinary creditors, including the CRA and labour. Allowing businesses to deduct interest expense when determining profit is consistent with commercial principles and is predictable and fair.

History of the interest deduction

O’Connor argues that Canada permits the deduction of interest expense because of legislative inertia. In reality, interest expense has long been viewed as an important cost of doing business. Almost all industrialized countries that levy an income tax (which includes most countries in the world) allow for the deduction of reasonable costs associated with earning revenue.

Generally speaking, reasonable interest expense incurred to gain or produce income has been deductible in computing income for tax purposes since 1923. Even in 1917, it was the position of Canada’s finance minister, Sir Thomas White, that

[w]ithout doubt, interest upon bonds is a fixed charge and net profits are only ascertained after deducting interest upon underlying charges, all interest payments, operating expenses and overhead—in other words, the net profits, according to a properly drawn balance sheet (Bryan Pontifex, Canadian Income Tax: The Income War Tax Act, 1917, With Explanations by the Minister of Finance and Instructions of Finance Department (Toronto: Carswell, 1918)).

When interest deductibility was first codified in the Income War Tax Act in 1923, then Finance Minister William Fielding argued that the amendments “do not introduce much that is new.” Rather, according to Fielding, they were “really intended to clarify the existing law.” An important point is that the finance minister proceeded to justify the deduction of interest as “a proper charge against the business” that “should be deducted before the income is declared” (Canada, House of Commons, Debates, June 27, 1923, 4492-4494).

Brian Arnold’s 1992 article in the Canadian Tax Journal, “Is Interest a Capital Expense?” addressed the question of whether interest is a capital expenditure that requires a specific carveout in section 20 for it to be deductible, or whether it is deductible in computing profit under section 9. Nonetheless, for a period of almost 100 years, interest has resembled any other deductible business expense and has been deductible in computing income from business.

Current law continues to permit taxpayers to deduct reasonable amounts of interest expense incurred to gain or produce income. Although the rules have introduced exceptions and limitations (for example, the after-tax financing rules, the foreign affiliate dumping rules, and the thin capitalization rules), this regime is longstanding, predictable, and fair. Interest is viewed as a cost of doing business, which is consistent with a normal understanding of how to measure business income.

Uncertainty and unfairness

O’Connor’s proposed change is radical. There would be winners and losers under the new regime, and it is difficult to predict whether the economic behaviour driven by the change would enhance Canada’s competitive advantage or harm it. Two aspects of the proposal, in particular, would add uncertainty and bring unintended consequences.

The first is that O’Connor’s proposal is not a universal denial of interest. He proposes a carveout, such that the prohibition would not apply to “moneylenders and businesses whose profits are, of necessity, a function of borrowed money.” It would be a complex task to determine which activities fall within these amorphous categories of business. Most credit providers are either in the business-to-business chain or in the business-to-consumer chain. Without a robust definition of O’Connor’s carveout—which, in our view, even O’Connor himself will find difficult to produce—Parliament will run the risk of reducing the availability of credit within Canada by substantially increasing the tax burden on businesses that provide such credit.

The second problematic aspect of O’Connor’s proposal is that many transactions involve elements that are not generally thought of as being interest at law but are akin to interest or have an embedded interest component. It will be difficult to determine which business expenses that involve a time value/interest component should be denied, and which should be allowed. Examples include rent, sharia-compliant financial products, repurchase agreements (repos), futures, options and forwards, discounts, and premiums. Even similar arrangements, such as early payment discounts on receivables, would result in an equivalent to interest payment. Under O’Connor’s proposal, it will be necessary to draw this distinction between the allowed and the disallowed, which will either add additional uncertainty or, as is likely, will default to purpose tests or to reliance on the general anti avoidance rule (GAAR), in order to distinguish taxpayers who are entering into equivalent-to-interest transactions for bona fide business purposes from taxpayers who are seeking to abuse the legislation.

Taxing income before interest is not equivalent to taxing “net income.” As Angelo Nikolakakis notes in this newsletter’s lead article, if tax is applied to earnings before interest and taxes (EBIT), a company may face a tax bill that is larger than its available cash. Taxing on a basis other than net income may therefore create distortions. Canada has a history of implementing, and later rejecting, non-income taxes designed to address perceived BEPS concerns. While many of these taxes did not endeavour to be revenue-neutral, they were eventually removed because “business taxes” that are not tied to net income (measured after ordinary business expenses including interest) distort the market. In the November 2020 Arnold Report, commenting on O’Connor’s proposal, Brian Arnold referenced the 1997 Report of the Technical Committee on Business Taxation. That report observed that Canada had responded to the erosion of its tax base by imposing other taxes, including capital taxes that impeded “prospects for job creation” (page 1.4). Arnold also referenced George Zodrow’s 2008 Policy Forum article in the Canadian Tax Journal, “Corporate Income Taxation in Canada,” which observed that foreign direct investment into Canada and international capital are “quite mobile and significantly affected by tax factors.” Canada has since discarded those taxes (including, for example, the tax on capital) as being unfair in that they were not tied to income or profit and did not promote competitiveness.

Undermining existing tax policies

Denying the deductibility of interest expense also signals that Parliament is opposed to borrowing to fund business activities. This signal will likely encourage businesses to underborrow. The ITA denies a number of deductions in order to deter behaviours that Parliament deems improper or excessive. Fines and penalties are non-deductible, even though courts have held them to be ordinary business expenses. Meals and entertainment expenses are generally limited to a 50 percent deduction. These express limitations signal to taxpayers that certain activity will not be underwritten by the tax system. Lumping interest into the category of non-deductible expenses suggests that incurring interest is undesirable and possibly “bad.”

Some of the reasons that O’Connor cites for eliminating interest deductibility run counter to other incentives that Canada has adopted to promote economic activity. Parliament has consistently taken the position that a reduced (or nil) rate of tax on an interest receipt is not a justification for taxing the payer.  For example, Canada has deliberately pursued zero rates of cross-border withholding on interest to promote the availability of low-cost credit to Canadian businesses. For many years, Canada imposed withholding tax on cross-border interest except where the so-called “5/25” exemption applied. No explicit rationale existed for the exemption, but it was commonly understood that this regime effectively prevented foreign lenders from competing with Canadian banks in the short-term lending market.

With the concurrence of the Canadian banks, the government deliberately changed the rules in 2007 in order to exempt all cross-border interest paid to arm’s-length lenders. The rationale for this change was to support Canadian businesses by facilitating cheaper and easier access to credit. O’Connor’s proposal would undermine this policy objective by subjecting cross-border interest to withholding tax as if it were a dividend. Parliament should not proceed with this change without re-evaluating the policy decision made in 2007, and we are aware of no reason to turn back the clock on that measure. The mere fact that Canada does not tax arm’s-length non-residents on cross-border interest in no way justifies a full denial of interest.

Another example is that deductions are also permitted for expenses paid to other taxpayers with sufficient losses to offset the income inclusion. There is no outcry that a Canadian business using its properly available carryforwards, such that it is not subject to tax, somehow subverts the tax system if another business claims an expense on amounts properly paid to the loss business for goods and services.

Parliament has also adopted the thin capitalization rules to preclude earnings stripping to certain related non-residents, seeking to strike a balance between the need of a Canadian business (usually a subsidiary of a non-resident) to obtain capital, and the base erosion that results from allowing an excessive interest deduction. In this regard, Parliament’s deliberate decision to limit the thin capitalization rules to cross-border loans from significant non-resident shareholders is an important signal regarding the amount of interest that will be considered reasonable to pay to certain related non-residents.

Not the right time

A cynic would argue that denying the interest expense deduction at this point in the economic cycle is politically expedient, given our low-interest-rate environment. The immediate impact on many businesses will be small. However, when interest rates eventually increase (and they will), the denial of interest deductibility has the potential to significantly increase tax revenues. This will not only disproportionately hurt businesses operating in capital-intensive industries, but also, potentially, further harm those businesses that were forced to borrow during the pandemic just when they are starting to get back on their feet. Given the pandemic’s ongoing unpredictability, these borrowings may be significantly larger than anticipated.

A cynic might also argue that it is mean-spirited to deny an interest expense deduction to businesses that have had to borrow during the pandemic in order to survive. Businesses rely on predictability, and during the pandemic, they would have made their current decisions based on a calculation of the cost of borrowing on an after-tax basis. To unexpectedly change the rules of the game in mid-stream would, at the least, reduce these businesses’ return on investment, and in some cases may jeopardize their survival. Under the existing rules, such borrowings would result in deductions and likely losses in the current year or two that would smooth out these businesses’ tax liability as they pick up and become profitable again.

Put yourself in the shoes of a business owner who is struggling to adapt to the pandemic and who now faces the regime proposed by O’Connor. You will probably see this radical tax change as ill-timed. You will likely have to spend more to invest in professional and other help to determine the impact of the new rules on your business. You will likely have to consider changing your business’s capitalization or renegotiating with lenders. You might also have to make changes in your business structure to separate financial businesses (which may be entitled to a deduction) from other businesses (which will not). If Canada needs to make tax changes now, we need to focus on making them simple (or specifically targeted), even if that means changing the headline corporate tax rate.

No one else is doing it

Tax competitiveness is also measured comparatively. Other countries allow full interest deductibility, so if Canada were to act unilaterally by adopting O’Connor’s proposal, the Canadian system would become comparatively unattractive for investment; businesses would face an additional expense when doing business in Canada. While some countries have changed their rules to limit interest deductibility to a maximum amount, these changes have been aimed at eliminating “excess” interest and reducing the incentives for businesses to “interest shop” internationally. For example, subject to certain exceptions, the OECD’s 2015 BEPS action 4 proposes a fixed cap on interest deductibility tied to EBITDA, along with targeted rules to address specific policy concerns. However, in targeting excessive interest—that is, interest in excess of what is reasonably necessary to earn income (see, for example, page 50 of the OECD’s Corporate Tax Statistics)—the OECD still recognizes that reasonable interest is a cost associated with earning revenue and should, accordingly, be deductible.


Back in 1996, in a paper titled “Why Tax Corporations?” Richard Bird identified many rationales for taxing corporations. Interestingly, he found no single rationale to be compelling, but he supported the taxation of corporations nonetheless, stating that “although none of the possible rationales for taxing corporations is particularly strong, in total it is clear that we not only should but must impose some explicit taxes on corporations.” He also observed, however, that such taxes should be properly designed “to collect revenue in ways that would improve economic well-being.” Denying an interest deduction to businesses in computing their taxable income is the wrong policy at the wrong time.

This article was originally published by The Canadian Tax Foundation in its publication titles "Perspectives on Tax Law and Policy".

To discuss these issues, please contact the author(s).

This publication is a general discussion of certain legal and related developments and should not be relied upon as legal advice. If you require legal advice, we would be pleased to discuss the issues in this publication with you, in the context of your particular circumstances.

For permission to republish this or any other publication, contact Janelle Weed.

© 2021 by Torys LLP.
All rights reserved.


Get in Touch