Empyrean’s Response to CRA’s Proposed Eligibility Guidelines

This post contains the feedback Empyrean submitted to CRA in response to their proposed draft SR&ED guidance document entitled “Policy on the Eligibility of Work for SR&ED Investment Tax Credits.”

 Scientific Method v. Engineering Method… “Hypothesis” v. “Prototype”

The added emphasis on the “scientific method” through this draft, as compared to the previous IC86, is plainly evident. Granted, the link between what is stated in the ITA (“systematic investigation”) and the “scientific method” was firmly established by Northwest Hydraulic Consultants Limited v. The Queen, so it is fully expected that eligibility for SR&ED should be defined in this context.

The Northwest ruling goes on to define the general steps of the “scientific method,” which are defined in this draft as requisites that differentiate standard practice and eligible SR&ED. In particular:

“A hypothesis must be clearly articulated and tested by experiment or analysis.”

In our opinion, the wording of this draft around the concepts of the “scientific method” and “hypotheses” will prove problematic.

This document fails to recognize that the “scientific method” is a generalized mode of systematic inquiry that may be used outside the realm of basic or applied science. The SR&ED program also includes experimental development, which relates to technology – the realm of engineers.

Engineers are not scientists, but their work can certainly result in technological advancement. Engineers do this by means of the “engineering method,” which is directly equivalent to the “scientific method,” as defined in this draft. Here is a comparison of the steps in both methods:

Scientific Method Engineering Method
State your question Define the problem
Do background research Do background research
Formulate your hypothesis, identify variables Propose alternative solutions; choose the best candidate and build a prototype
Design experiment, establish procedure Design experiment, establish procedure
Test your hypothesis by doing an experiment Test your  prototype by doing an experiment
Analyze your results and draw conclusions Test and redesign as necessary
Communicate results Communicate results


In an engineering context, “articulation of a hypothesis” is identical to “building a prototype.” The “hypothesis” is implicit in the design of the prototype. (Here, “prototype” is used in its general sense.) A prototype embodies a proposed means to overcome a gap in technological knowledge (an uncertainty). In an engineering context, the act of proposing, building and testing a prototype is exactly equivalent to “articulating a hypothesis.”

This is the reason one almost never finds a “hypothesis statement” in the engineering world, but almost always finds “a prototype design.”

A prototype is the “articulation” of a hypothesis.

(It also defies logic that a company would allow its engineering team to invest the time and cost to propose, design, build and test a prototype unless it there is a significant technological uncertainty to overcome — if not, why wouldn’t it go straight to the answer by applying standard practices?)

The language in this draft, however, does not recognize this fact. It is so narrow that the real potential exists to exclude otherwise eligible work based on the fact that an engineer failed to act sufficiently like a scientist to CRA’s liking.

In Northwest Hydraulic Consultants Limited v. The Queen, Judge Bowman alludes to the lopsided power inherent in the wording of any CRA guidance document:

In general, I am reluctant to rely too heavily on interpretation bulletins and information circulars in determining contested issues under the Income Tax Act. The reason for this is that in any litigious situation it seems somewhat unfair for an independent arbiter to place much weight on the rules of the game devised by one of the players.

Moreover, Bowman goes on to imply that rather than using a strict adherence to one definition of the “scientific method,” that definition should be elastic enough to include

intuition, creativity and sometimes genius in uncovering, recognizing and resolving the new uncertainties.

Here is another example where a strict adherence to the draft’s definition of the “scientific method” fails. The T4088 defines “Mathematics” (Code 1.01) as an admissible Field of Science or Technology. However, it is an ongoing debate among experts as to whether Mathematics should even be classified as a “science” (it is certainly not a technology). Furthermore, advancements in mathematics are almost always achieved without using the scientific method (e.g., inductive or deductive reasoning).

So, where does that leave this draft’s more dogmatic definition of the “scientific method?”

In our opinion, this draft fails to capture Judge Bowman’s intent by relying so heavily on a science-oriented definition of the “scientific method.” This aspect of the draft also conflicts with the need to consider a company’s business environment — i.e., one in which engineers, not scientists, are used to develop new and improved products and processes.

Or is it CRA’s intention to say that the only legitimate SR&ED must be carried out by scientists?

Surely not.

Removal of Reference to “Slight” Advancements

The existing IC86 reflects Judge Bowman’s intent regarding the need to consider even “slight” advancements:

I should not have thought it was necessary to say so. Most scientific research involves gradual, indeed infinitesimal, progress. Spectacular breakthroughs are rare and make up a very small part of the results of SRED in Canada.

However, the proposed draft removes any reference to “slight” improvements. Is this to be interpreted that CRA will no longer consider “slight” advancements eligible? If so, is it CRA’s intent to admit only “spectacular breakthroughs?”

Broadening of the Definition of “Standard Practice” and “Routine Engineering”

The language around “standard practice” and “routine engineering” in this draft is substantially expanded compared to the existing IC86. In our opinion, it broadens the scope for what could be interpreted by an RTA to be “routine engineering.”

Particularly troubling are statements like this one:

Standard practice refers to directly adapting a known engineering or technological practice to a new situation when there is a high degree of certainty that the known technology or practice will achieve the desired objective.

Here are some ambiguities that may prove problematic:

-          Who defines “a high degree of certainty?” Is it a team of company engineers with decades of experience, or an RTA who has no in-depth knowledge of the nuances of the problem at hand? Too often, an RTA can make sweeping assessments based on only a cursory assessment. This language only empowers this unreasonable process.

-          What if the project starts out with a “high degree of certainty” only to lead to failure? Judge Bowman recognized that projects can involve “…uncovering, recognizing and resolving new uncertainties.” The existing IC86 opens the door wider to this possibility, while the new draft seems to slam it shut.

-          Equating “standard practice” to the “directly adopting known engineering or technological practice” can lead to the all-too-common phenomenon where an RTA breaks down a project’s activities to the “atomic level” where everything appears to be “known engineering or technological practice.” Example: “You performed an FEA to analyze the prototype, but FEAs are standard engineering practice, therefore it is ineligible.” This specious reasoning may be encouraged by the language of the new draft because it fails to balance the admonitions about “standard practice” with a reminder that an SR&ED project must be considered at the level of “interrelated activities” and not broken down ad absurdum.

Respectfully submitted,


Dr. Frank Naccarato, P.Eng.

President, The Empyrean Group Inc.

Comments are closed.