The Myth of Canadian Industry’s Poor R&D Performance
Canada’s R&D Expenditures have been underreported for quite some time. Numerous reports on Canada’s innovation economy have noted Canada’s poor record of private sector spending on research and development (R&D) relative to other Organization for Economic Cooperation and Development (OECD) countries. While we are among the middle of the pack on public R&D spending, we rank near the bottom when measuring Canada’s R&D expenditures (BERD).
However, these statistics may all be incorrect. Because of the way that the data has been collected, it appears that Canada’s BERD may have been significantly underreported over an extended period of time.
Until very recently, annual spending on R&D by Canadian businesses has been reported only for work that meets the definition of Scientific Research and Experimental Development (SR&ED) in the Income Tax Act. The definition of R&D used by most OECD countries is substantially different; it uses the current Frascati model that includes R&D in the humanities and the social sciences, and product development based on existing knowledge.
Because of these differences in definitions, Canada may have been underreporting its expenditures on R&D since 1997. It is difficult to estimate the level of underreporting spread out over nearly two decades, but we tried to gauge the extent of the problem by looking at salaries spent on R&D. We tabulated the differences between the total salaries spent on personnel involved in R&D activities and the amount allowed to be claimed for SRED purposes for a sample of ten companies. Although measuring total R&D salaries will overreport the amount that would be allowed under the Frascati definition, it is indicative of the problem that the differences in definitions have created. Our conclusion is that the R&D for our sample of companies may be underreported by up to 73%.
Numerous reports on Canada’s innovation economy have pointed out not only that Canada’s BERD expenditures are lower than the OECD average but also that they appear to be declining over time.
In recent years, the administration of the SR&ED program has become more structured, and expectations have been set for private sector R&D work to include formal experimental procedures and complete evidence records. According to program users, it is more difficult to get SR&ED credits approved now than in the past.
This may mean that not all R&D that was reported by companies in past years is captured with the tightening of the SR&ED program. Hence, if SR&ED is the number used by Statistics Canada for R&D reporting nationally, then R&D expenditures would have declined as the program was tightened.
But if we are using an incomplete definition of R&D that leads to erroneous conclusions, then have we been developing flawed R&D policies and expenditure programs?
And if our R&D spending has been higher than reported, why aren’t we doing better?
Unfortunately, without a substantial amount of work, it is unlikely that we will ever know how much business has actually been spending on R&D. Thus we may have lost critical time that could have been productively used to develop a more targeted innovation policy. We can only hope that recent changes to data collection by Statistics Canada will enable us to rectify these problems.