Ron Freedman, CEO, Research Infosource Inc.

Reform SR&ED by adopting TRL model

By Ron Freedman

Most analysts are dismayed by Canada's fair-to-middling performance in the international R&D sweepstakes. According to the OECD, our GERD/GDP ratio (total national spending on research as a percent of GDP) is 1.62%, which puts us in about 20th position among advanced economies. Nearly as bad is the business sector's research performance, where we rank 14th overall.

Over the years many reasons have been put forward for this lacklustre situation: industry sector composition/structural economic factors; fewer large companies; small markets/lack of scale; low competitive intensity; shortage of venture capital funding; poor commercialization of university research; shortage of direct government funding for research/over-reliance on indirect funding; poor business innovation strategies; and, the catch-all multi-factor productivity deficiencies. Indeed each plays a part.

Contributing factors fall into two categories — those under the control of businesses and those under the control of governments. With regard to the latter, policies and programs concerned with the funding of business innovation must rank at the top. And the program that has survived decades of calls for improvement is the Scientific Research and Experimental Development (SR&ED) program, which offers investment tax credits totalling around $3.5 billion annually. Based on the evidence, if the policy objective is to boost business spending on research and innovation, SR&ED simply isn't working.

About SR&ED

The SR&ED program is the cornerstone of federal government support for business research. In addition, a number of provincial governments have programs that top-up federal funding. How does SR&ED work? SR&ED will fund basic and applied research and experimental development, defined as:

Basic research to advance scientific knowledge without a specific practical application in view, usually carried out in a lab setting. Applied research to advance scientific knowledge, but with a specific practical application in view. Experimental development to achieve technological advancement for creating new, or improving existing, materials, devices, products, or processes, including incremental improvements.

Unexplained, and utterly illogical, is why government would encourage companies to engage in research with no practical application in mind, let alone fund them to do so?

Companies that engage in applied research that is not hypothesis driven and capable of being published in scientific journals are not eligible for SR&ED support. So strictly speaking, world-changing innovations such as the steam engine, electricity and telephone would not have qualified for SR&ED. Does this make sense?

Research in the Real World

The problem is, with some exceptions, innovation in the real world of business seldom works like the academic paradigm suggests it should. Most companies don't begin research in a laboratory and after years of publishing results in scientific journals, roll out products. Most identify customer needs and assemble existing technologies (some their own and some borrowed) into new products or processes. Occasionally, they will develop technologies themselves; but almost never by starting with basic research.

The exceptions to the hypothesis rule occur largely in the life sciences (i.e. pharmaceutical, genomics) and a few large firms such as IBM that still conduct basic research in their own laboratories. In addition, a small number of companies pull early-stage research out of universities and government labs and develop them into new products. But they seldom do the basic hypothesis-driven work themselves; it's too risky, time-consuming and expensive.

Why then, does the Canada Revenue Agency persist in using an outmoded paradigm of technological change to fund business research when the real world has clearly moved on? Much of the Canadian position comes from slavish adherence to international (i.e. OECD) definitions of research that were developed in the 1960s and still canonize the laboratory-to-market paradigm. That plus a reluctance to let go of the failed orthodoxy because that would be an admission of error.

Another cause is the failure to decide on what it is that governments want firms to do in the first place. The answer is quite simple; we want firms to develop tradable goods and services. The focus of public policy should be on providing firms with the technological capacity needed to develop products and services before they enter the marketplace.

In reality, policymakers should (with the above exceptions) be dissuading companies from engaging in basic and hypothesis-driven research and leave that noble activity to academics and in some cases government labs. Otherwise companies will be squandering internal resources far too early in the product development cycle.

The current CRA approach has un-intended consequences. Firms bend themselves into pretzels to pretend that their R&D is hypothesis-based. They are forced to hire legions of costly SR&ED consultants (often ex-CRA employees) to make it seem thus to CRA auditors. In the process, companies are motivated to spend too long and too much on basic research — because that's where the money is in the short term — and too little on developing capability.

A Better Way

Is there a better approach? You bet. The SR&ED system should ditch its current hypothesis fixation and adopt the military's nine-stage Technology Readiness Level approach to supporting research and innovation. SR&ED should be used to fund companies at TRLs 1-6 (basic research to prototype development), which are the steps required to develop technological capability. Levels 7-9 cover standard product engineering and as such should not be funded. Note that the current hypothesis-based approach is still covered in TRLs 1-2, so companies working at those stages can still receive tax credits.

The key test that should be applied to individual projects is "Does this project significantly improve the technological capabilities of the firm?" If it does, for example by introducing a new or substantially different technology into the firm, then it should qualify. Simple. Hypotheses if necessary, but not necessarily hypotheses.

In summary, it's time to ditch the outmoded SR&ED definitions and reform the program by adopting a TRL-based approach.

Ron Freedman is CEO, Research Infosource Inc. ron@researchinfosource.com.

(A longer version is available online).