Innovation has become the holy grail of economic policy-makers in Canada. There is nearly universal agreement among economists that Canada needs much more of it to ensure the long-term prosperity of its citizens, the sustainability of its public finances and the competitiveness of its exporters. The broader public is also generally in favour of innovation but balks at the notion of productivity, erroneously believing that it means requiring employees to work harder or investing in machines to make them obsolete. (Once they understand that productivity involves working smarter, stronger wage growth and more business opportunities for the companies they work for, their views typically change.)

There is far less agreement on what innovation actually is, how it is created and operationalized, and to what extent the government can or cannot encourage its creation and adoption. Canada has spent the better part of two decades on countless expert panels trying to get a handle on what causes innovation, why Canada doesn’t do it as well as other nations and what remedies are available.

The fundamental problem is that discussions of innovation in Canada inevitably fall into the trap of equating it with research and development, science and technology, based on the following line of reasoning:

  • Investments in science and R&D are key ingredients to innovation.

  • Innovation leads to higher productivity (by either reducing the cost of producing existing goods and services or creating new higher-value-added products and services, or both).

  • Therefore, Canada must invest more in R&D to improve innovation and productivity.

As with all logical fallacies, both premises are indeed true, having been confirmed by decades of empirical economic research in Canada and around the world. But the conclusion depends on an unstated assumption that the supply of science and R&D creates its own demand for innovation and the enhanced productivity it brings: create the knowledge and the innovators will come.

But this unstated assumption, which has formed the basis for the ”œscience-push” innovation policy in Canada for at least the past 40 years, is simply not true, for if it were, we would not have a need for a special issue of Policy Options on the subject nor the high-profile symposium being held in conjunction with its release. For Canada has significantly boosted direct funding for university R&D in the natural and social sciences, has one of the most generous tax subsidies for business R&D and is generally not lacking in knowledge and scientific know-how.

Figure 1 demonstrates the failure of science-push innovation policy by plotting key measures of R&D and science funding against productivity since 1997. What is most striking about the figure is that federal investment in science and technology (S&T) " the backbone of conventional innovation policy " rose by 60 percent from 1997 to 2009, while business productivity grew by a paltry 16 percent. In addition to this large gap, the timing in terms of productivity dynamics is the opposite of what a believer in science-push policy would expect. The bulk of the increase in S&T investment occurred between 1997 and 2001, and since the fruits of the new knowledge those investments created would have taken some time to cultivate and harvest, one would have expected an uptick in productivity starting in the 2000s. But the uptick, small though it was, occurred contemporaneously and was, as verified by empirical research, due mainly to the impact of the rise of the World Wide Web and the opportunities it provided for offering new services and cutting costs. Indeed, the rise in business R&D during the 1997-2001 period (which was stimulated in part by such Web opportunities) was the more important factor. Since 2000, productivity has essentially flatlined.

Despite the clear lack of correlation between R&D (whether financed by business or government) and productivity, debates about innovation remain stubbornly centred on research and development. Part of the reason is that R&D is one of the few metrics for which detailed and reliable internationally comparable data exist, thus making it easy to see how Canada stacks up against its international peers. The data in support of this notion have been charted countless times, and figure 2 adds to this collection by showing total R&D as a percentage of GDP broken down into its three main components: business enterprise R&D (BERD), higher education R&D (HERD) and governmentperformed R&D (GOVERD). Data are shown for the G7 nations as well as two other countries that are more economically similar to Canada in that they have relatively large natural resource sectors: Australia and Norway.

In terms of total R&D, Canada devoted 1.86 percent of GDP to R&D spending in 2008 (almost exactly the same share as in 2000), which placed it fifth in the G7 and second among the three resource-intensive countries.

Let’s focus first on BERD. Within the G7 Canada’s ranking falls to sixth (ahead of only Italy), and it is well behind Australia and slightly ahead of Norway. It is easy to see from figure 2 that Canada devotes the thirdlowest percentage of its GDP to business R&D of the nine countries shown, at 0.99 percent. This is less than half the share of the United States, the seemingly natural point of comparison, but closer to the other resource-rich nations in the group. This relatively poor performance is all the more ironic because Canada has one of the most generous R&D tax credits among industrialized nations. It provides a subsidy of 20 cents on the first dollar of research (35 cents for small businesses), and thus certainly subsidizes a large amount of research that would have been undertaken in any event.

In addition, sectoral analysis of the distribution of business R&D shows that the gap with the United States is not a function of intensity per se; it exists because research-intensive industries in Canada are much smaller than their US counterparts. For example, research intensity in the Canadian pharmaceutical and computer equipment industries is actually higher than in the United States, but these sectors are much smaller than those south of the border.

Finally, because the government has no way of monitoring the effectiveness of business R&D in stimulating innovation and productivity, it is impossible to know the net benefits to the taxpayers who fund it. A March 2011 investigative report by the Globe and Mail suggests that more than one-third of the $4.7 billion in R&D tax breaks dispensed in Canada is being wasted as a result of vague rules on what expenditures qualify for the credit, indicating that it is walking a fine line between effective policy instrument and boondoggle.

Canada ranks far more favourably with regard to HERD " the R&D that is performed in universities. In 2008, Canada devoted 0.68 percent of GDP to university research, the highest of the eight other countries. This share has increased significantly since 2000. In sharp contrast, the United States, Japan and Germany, widely regarded as the world’s leaders in innovation, spend just one-third to one-half of this share on university R&D.

This anomaly provides prima facie evidence of a large divide between university research and industrial innovation. At every innovation conference I attend, I ask the CEOs of Canadian companies talking about their success es in becoming innovative the extent to which university research was an important factor in their success, and the answer is almost uniformly ”œLittle to none.” This admittedly anecdotal evidence is corroborated by more rigorous, though dated, survey results: among manufacturing plants that indicate having made innovations in the 2002-04 period, in most industries (with the exception of pharmaceuticals, base metals, and plastic and rubber products) less than 5 percent reported licensing technologies from universities.

But one of the main motivations for boosting funding for university research is to enhance productivity. Knowledge for its own sake is a noble undertaking, but the new funds invested in Canada since 1997 were clearly intended to have an eventual economic impact. Consider an excerpt from Paul Martin’s 1997 budget speech: ”œThe research facilities in our hospitals, our universities and our colleges are part of the root system of our economic prospects for the future.” From this perspective, it can be regarded only as a policy failure that so much investment in university research apparently has yielded so little in terms of practical application in business.

As the IRPP’s Jorge Niosi explained, the fundamental challenge of bridging the gap between knowledge generated in universities and the business needs of industry is that the existing technology transfer infrastructure is ill-suited for its intended goal. Thousands of papers are published in hundreds of learned journals, and most university technology transfer offices are staffed by just a handful of individuals who cannot be expected to keep abreast of all potential markets for the ideas that come across their desks. Furthermore, the academic background of the staff means that they may not have the relevant experience and knowledge to assess the commercial value of those ideas.

A more effective way of harnessing the potential business value of university research is to turn the existing science-push model on its head. Businesses are much better equipped to judge the commercial benefits of university research, and they should thus be more present in the administration of government programs to promote technology transfer.

The US provides a good blueprint with its Small Business Technology Transfer (STTR) program, which for almost 20 years has provided grants for partnerships between universities and small businesses. But the grants are firm-driven: the business seeks out university researchers doing promising work and communicates clearly the lines of investigation that have the most commercial potential. There is even a legislative requirement that the researcher work at least half time at the sponsoring firm. When more decisionmaking power is put in the hands of business, the innovation payoff from the research is much larger. Business involvement in university research, long considered taboo, must be increased in the context of clearly circumscribed programs if Canada is to reap the benefits of being a leader in higher education R&D.

But significantly increasing the innovation ”œthroughput” of university research is not enough. The fundamental observation that follows from the fallacy of R&D push leading to innovation and productivity is that other factors besides R&D are necessary for knowledge to lead to innovation. Research by Industry Canada conducted in the early 2000s modelled innovation as a composite of four factors: R&D, the level of patenting activity (to measure the business value of inventions), the educational attainment of the workforce (to measure the ability of employees to operationalize innovation) and investment in machinery and equipment (the vehicle through which many new technologies reach the marketplace). R&D spending alone has no statistical relationship with productivity, while the composite index of all four factors has a significant positive relationship.

Finally, and perhaps most importantly, we have to remember that innovation is much more (or, to speak more accurately, much less) than science and has more to do with competitive pressures in the marketplace. Consider Federal Express, which redefined the pace of business activity with the insight that there was a latent demand for next-day-guaranteed delivery of letters and packages. This idea did not require massive amounts of cutting-edge academic research, but was instead a textbook example of how competitive marketplace pressures (in this case, driven by customers) can lead to non-science-based innovation. To cite the timeless adage, necessity is the mother of innovation.

Competitive pressures have a central role in contemporary economic models of innovation, and research done at the OECD shows fairly conclusively that anticompetitive regulation of industries inhibits investment in information and communications technologies, a key vehicle for bringing innovation to the marketplace. They find that ICT-intensive industries in Canada tend to have less-competitive markets than in 11 other OECD countries, including the United States and Australia. This goes a long way to explain why Canada invests only about half as much per worker in ICT equipment as the United States.

Other factors affecting the competitive environment are well known and have been studied by numerous groups, most recently the Competition Policy Review Panel chaired by Red Wilson. They include barriers to internal trade, foreign ownership rules and other barriers to entry in certain large sectors for which innovation is an important potential growth source. While a full treatment of these issues is not possible here, the fundamental point is that the demand side of innovation needs much more policy attention than the supply side.

In his assessment, Donald McFetridge concluded that Canada’s innovation policy is at a strategic crossroads and could take two broad directions. One departs from the science-push approach of the past several decades (which has clearly not delivered on promises of increasing productivity growth) and would focus government efforts on removing obstacles to competition in whatever form they may come. This would stimulate demand for business R&D, university knowledge and all of the other sources of good ideas (of which there is no shortage of supply in Canada). The other, more politically expedient path would continue the current practice of paying lip service to the demand side of innovation but perpetuate the fallacy that patchworks of science and technology incentives can push ideas downstream to a marketplace that may not need or want them. If Canada truly wants to become an innovation nation, it would do well to choose the former over the latter.