-A A +A

Finance Minister Chrystia Freeland tables the federal budget Thursday. She should use the occasion to implement policies that would improve Canada’s lacklustre innovation. A useful step in that direction would be to announce that income from patents and other intellectual property (IP) will be taxed at a special low rate. Generally described as an “IP Box,” such an initiative would boost Canada’s flagging R&D spending, raise our low commercialization rate, and stem an outflow of IP profits to tax havens.

As I explain in a C.D. Howe Institute paper published today, two developments make this a good time for an IP Box. The first is an OECD-inspired international agreement that income taxed at a preferential rate must be derived from R&D actually performed in the implementing jurisdiction. With that’s true, preferential IP regimes complement support for R&D provided through investment tax credits and defend against poaching by tax havens without igniting a lose-lose competition for mobile profits.

The second development is the tentative international agreement to impose a 15 per cent minimum corporate income tax rate on large multinational enterprises (MNEs). If enough countries end up joining, aligning the IP Box rate with the global minimum would eliminate the incentive for large MNEs to book IP income in tax havens. That would substantially reduce the fiscal cost of implementing an IP Box. Without a global minimum tax rate, Canada would continue to compete with low-tax jurisdictions such as Barbados, which has a one per cent tax rate on corporate income exceeding about C$20 million. Competing with Barbados would reduce, not increase, Canada’s tax revenue from IP income.

A preferential IP regime would have to be carefully designed to reap all its potential benefits. The key design feature is to define qualifying IP income as broadly as permitted so that it captures income from all IP developed from R&D performed in Canada, including IP protected by patents, trade secrecy, and software copyright. Restricting the benefit to income from patents would bias R&D spending towards inventions, which are typically patented, but account for well under half of all successful R&D projects.

To minimize administration and compliance costs and prevent tax-induced changes in the location of R&D across the country, a preferential IP regime should be implemented at the federal level. And it should be a deduction from taxable income instead of a tax credit. A deduction would reduce the tax bases of provinces that have tax collection agreements with the federal government, effectively ensuring they share in the costs as well as the benefits of the program. (They share in the benefits when companies book more R&D in their province, thus raising their tax base.)

Participation by Quebec and Alberta, which have not signed tax collection agreements, would be at their discretion. Québec already has a preferential IP regime. A federal deduction set to align the federal-provincial tax rate on IP income in tax-collection provinces with a 15 per cent global minimum tax rate would push the IP Box rate in Québec down to 10-11 per cent. This is likely below the revenue-maximizing rate, so Quebec would have an incentive to raise its preferential IP rate.

A carefully designed IP Box is likely to be more cost-effective — i.e., produce a larger impact on R&D per dollar of tax revenue forgone — than an equivalent increase in the regular Scientific Research and Experimental Development (SR&ED) investment tax credit. Although the SR&ED tax credit is paid up- front while an IP Box rewards successful projects after the fact, each increases the expected after-tax return to R&D by the same amount per dollar of tax revenue forgone, when adjusted for the time value of money. As a result, the investment response of firms that can borrow on favourable terms is likely to be similar for both measures. But the net loss in tax revenue from an IP Box would be less because less income would be booked in tax havens, thus tipping the cost-effectiveness ratio in its favour despite what would likely be higher administration and compliance costs.

An IP Box aligned with a 15 per cent global minimum tax rate would be equivalent to increasing the regular SR&ED tax credit rate by almost three percentage points. But overall support for R&D performed by large firms would still be well below the level that would maximizes its benefit to Canadians. Maintaining the SR&ED tax credit at its current level when implementing an IP Box would therefore be good public policy. The end result would be more R&D performed and commercialized in Canada, financed in part by reduced profit-shifting with tax havens.

John Lester in an Executive Fellow at the School of Public Policy, University of Calgary.

Published in the Financial Post