Standard Practice Software Development: Hypercube Inc vs. Her Majesty The Queen

Hypercube Inc vs. Her Majesty The Queen demonstrates how the CRA apply standard practice or routine engineering to deny weak software development claims.


I’ve broken down the case by the Five Criteria established by Judge Bowman in the case of Northwest Hydraulic Consultants Ltd. v. Canada which is commonly referred to by the Tax Court.

1. Was there a technological risk or uncertainty which could not be removed by routine engineering or standard procedures? 

The project,

 consisted of developing a program to read and analyze source code from Web sites to detect weaknesses

This was to be a web crawler designed to replace a manual process of analyzing website code and provide a report to help programmers do their checks.

The Judge referenced the facts:

[18] Mr. Villeneuve noted that no existing technology allowed this to be done. Current technology, he explained, could not collect the information, validate it and retrieve it in the way they wanted it to. Mr. Villeneuve stated in his testimony that the program would improve the underlying technology. He described the underlying technology as being various programming languages.

[21] Regarding the technological uncertainties facing the appellant, Mr. Villeneuve stated that they were connected with the underlying technology. The uncertainties related to whether the underlying technology could be improved so that a crawler could be programmed as required to carry out the project.

Available languages may be a limitation but Improving programming languages seems to be beyond the scope of creating a new web crawler. Additionally, it is contradicted by an earlier statement that suggest the project started when the required tools were available :

[11] Mr. Villeneuve stated that he began thinking about this project in 2010 and discussed it with the appellant’s other employees around the beginning of 2012. Page: 3 The project was the product of a reconsideration of the various programming tools available.

2. Did the person claiming to be doing SRED formulate hypotheses specifically aimed at reducing or eliminating that technological uncertainty?

The hypothesis established by the claimant was that,

if a new way of performing Internet diagnostics was adopted, significant effects in terms of technological advancement could be observed”.

This isn’t a scientific hypothesis as it does not propose a “new way” and does not postulate an effect.

3. Did the procedure adopted accord with the total discipline of the scientific method including the formulation, testing and modification of hypotheses?

The work described failed to demonstrate any modifications of the hypothesis:

[20] Mr. Villeneuve explained that the project was carried out methodically, using a process of [TRANSLATION] “trial and error”. The appellant’s employees analyzed whether they could retrieve an initial piece of information in a Web site’s code, validated their attempt and then made another attempt to obtain a second piece of information, and so on.

[22] Mr. Villeneuve then spoke in more detail about what he saw as the problems they experienced in developing the program and explained the various measures taken to resolve them.

[23] An initial problem was that the program did not work for certain domain names that were being used by the Web sites being analyzed. Changes were made to the program so that it would work with all existing domain names.

By explaining the work as a series of problems that were resolved individually only standard programming work was demonstrated.

4. Did the process result in a technological advancement?

While they succeeded in creating a new website analysis tool, the judge applied the following logic:

[47] Although the appellant’s program could constitute an entirely new product, it was created using well-known techniques. Novelty or innovation in a product is not sufficient to illustrate technological advancement.

The claimant failed to demonstrate any new knowledge gained from their work.

5. Was a detailed record of the hypotheses tested, and results kept as the work progressed?

There was little documentation submitted to the court and, as the judge notes, the evidence was not explained.

[48] Moreover, the appellant produced very little evidence documenting its project. The only documents introduced in evidence are the program’s tree diagram and a log of hours worked. This tree diagram was not specifically explained in Court, and the time log does not appear to reflect reality. I do not think that this evidence is sufficient to support an SR&ED claim as prescribed by the ITA.

Documentary evidence could have helped verify the testimony but the claimant had already failed to meet the other criteria.


This case demonstrates poor project framing. The objective was a high level advancement in the “underlying technology”, namely programming languages, but the work was low level resolution of issues. I think the project could have qualified if described somewhere in the middle.  A more feasible advancement would have been an improvement web crawling methods; the underlying technology behind their objective to automate website analysis.  The hypothesis would build on this by stating the web crawling methods they thought could improve the crawlers performance. The experimentation would then focus on more fundamental changes to the web crawling methods and associated variables with any issues encountered classified as support work.


Do you think this project should have qualified for ITC’s? If so, how would you frame it?


The Riskiest Industries to Claim SR&ED

Every SR&ED claim carries risk of a CRA review. But what are the chances of getting the amount claimed after a visit from the CRA? Only 16% of claims were reduced following an audit from 2007-2009; however, some industries had much higher rates. Information revealed in the Muller Report  shows the disallowance ratios of audited claims (for industries with over 100 claims from )

Industry # Claims Ratio
Heavy and Civil Engineering Construction 100 80%
Furniture and Related Product Manufacturing 246 39%
Building Material and Supplies Wholesaler-Distributors 163 38%
Clothing Manufacturing 143 36%
Wood Product Manufacturing 262 36%
Management of Companies and Enterprises 136 35%
Primary Metal Manufacturing 112 30%
Paper Manufacturing 143 29%
Crop Production 336 28%
Animal Production 270 28%
Miscellaneous Wholesales-Distributors 223 27%
Food, Beverage and Tobacco Wholesaler-Distributors 136 27%
Specialty Trade Contractors 326 25%
Food Manufacturing 615 24%
Fabricated Metal Product Manufacturing 908 24%
Electrical Equipment, Appliance and Component Manufacturing 235 22%
Plastics and Rubber Products Manufacturing 440 18%
Support Activities for Mining and Oil and Gas Extraction 106 17%
Repair and Maintenance 263 17%
Overall Average 16%

Almost half of the industries in the list are manufacturing with typically low innovation categories such as clothing, furniture and primary metals near the top and higher tech categories such as electrical and plastics near the bottom. It has been well known in the industry that SR&ED does not favour manufacturing and this data confirms it.

Agriculture is another high risk industry. This probably stems from the difficulty of separating experimental production from commercial in an agricultural environment. The boundaries need to be clearly defined; experimentation entails risk and the CRA does not believe someone would risk their entire crop for an experiment.

Other industries that relate to activities specifically exempt from SR&ED also appear on the list, such as wholesalers, management and oil & gas. Subsection 248(1) of The Income Tax Act states that SR&ED “does not include work with respect to market research or sales promotion, research in the social sciences or the humanities, prospecting, exploring or drilling for, or producing, minerals, petroleum or natural gas.  Any activities falling into these categories are easy targets for the CRA. Successful claims tend to fall outside of the companies primary focus and actually fall into sciences such as software, mechanical engineering or robotics.

The highest ranked industry is heavy and civil engineering construction with over twice the disallowance rate of any other industry. This could be partly due to similar reasons as agriculture; the separation of commercial from experimental production but the challenge in construction goes beyond that. Its about risk. Typically the size of projects in this industry mean that only one is built and once that project is built it cannot  fail. In terms of SR&ED this means all the prototyping is done in simulation or modelling which the CRA can classify as due diligence and there are few failures to prove uncertainty.

This is one example of the insightful data available in the Muller Report. Sadly this information is from 2009. Hopefully the CRA will become more open and share data so we can all work to improve the program.

CRA’s Risk Factors for SR&ED claims

The Canada Revenue Agency’s Claim Review Manual for Research & Technology Advisors (RTAs) includes the following risk factors used for selecting SR&ED claims for review:

  • filing history of the claimant;
  • claimant’s compliance history;
  • work described in the project descriptions;
  • sector-specific issues;
  • nature and extent of the non-compliance issue;
  • amount of ITC at risk;
  • opportunity to educate claimant on the requirements of the SR&ED Program and to inform claimant about available claimant services and literature;
  • opportunity to promote self-compliance and proper self-assessment by the claimant;
  • opportunity to obtain a better understanding of the claimant’s work; and
  • effect of either carrying out a review or not carrying out a review on future compliance.

Once selected, the following factors are used to identify high risk projects/activities:

  • Large projects/claims in relation to the industry sector or the claimant’s norm or resources;
  • Projects that seem unusual for the claimant’s business line;
  • Project descriptions that strongly suggest that work is not SR&ED or where the descriptions make it difficult to determine what was done;
  • Indications of lack of separation of SR&ED and commercial activities, particularly where the claim may involve questions of Experimental Production (EP) or Commercial Production with Experimental Development (CP+ED);
  • Projects that are ongoing for many years with no clear end in sight;
  • Historically significant problems with the issue, with the claimant or with other claimants that had similar issues;
  • Issue could affect other issues or other companies with the same issue;
  • The materiality of the issue;
  • Work may have resulted in the creation of a significant commercial asset;
  • Known R&D centres outside Canada; and
  • Issue is material, typically the amount of ITC at risk relative to the total claim.



Pixar’s Rules to Phenomenal SR&ED Reports

I never realized how much writing technical report for SR&ED projects is classic story telling but many of Pixar’s 22 Rules to Phenomenal Storytelling match up well to the SR&ED narrative. For example:

1. You admire a character more for their trying than their success.

In SR&ED terms, this means focus on the project’s failures and the efforts to overcome them instead of just listing accomplishments.

4. Once upon a time there was ____. Every day, _____. One day ____. Because of that, ____. Until finally ____.

This is the SR&ED project structure, for example: Once upon a time there was a company trying to improve their process (uncertainty). Every day they ran the machines the same (standard practice)  One day they experimented with new variables (hypothesis/work done). Because of that they improved production (result). Until finally they realized how the variables affected their process (technological advancement).

7.  Come up with your ending before you figure out your middle. Seriously. Endings are HARD, get yours working up front.

For SR&ED the ending is the advancement. This suggests defining your advancement first then working backwards from their. A great tip that I am going to try applying.

Can you find any other similarities? Does this apply to other industries such as sales. Post a comment and let me know.



What is SR&ED?

By the Numbers:

  • $3.6 billion credits from the federal government and $1 billion from the provinces annually makes SR&ED Canada’s largest business tax credit (1)
  • 18,000 claimants a year
  • 35% of eligible expenses refunded to Canadian Controlled Private Corporations (CCPC) federally (2) with provinces adding anywhere from 10% to 37.5% (3)
  • 18 month deadline from the fiscal year end (4)
  • 120 or 240 day turnaround times (depending on when you file) (5)

What Qualifies?

Any actives undertaken to improve products and/or processes using the scientific method. This can include:

  • Complicated issues that required experimentation to solve
  • Incremental improvements to product/processes
  • Even failed attempts as long as you learned something (6)

You can then claim related :

  • Employee hours (with an overhead bonus)
  • Materials consumed in testing
  • Subcontractors (at a reduced rate) (7)

How To Apply?

To apply, file a T661 form (8) with your corporate tax return. While I have simplified things here, the program is very complicated and always changing (see the citations linked to CRA websites).  If you have any questions fill out the form below and I’ll get back to you.

Some thoughts on the CRA’s SR&ED draft examples

While examining the CRA’s newly release SR&ED examples, I noticed a couple of trends in the content. I’ve attempted to relate these finding to the CRA’s new review approach. Here is what I found:


Technology Reviewers commonly site this as a reason to deny a project.  A report stating that work is standard engineering and not SR&ED can be hard to dispute. The examples reiterate that this strategy will continue. Although only Example 4 explicitly states the example “shows standard practice”, the first 5 (of the 10 examples) relate to standard practice.

How do you prove that your objective could not be accomplished through the use of standard methods or available technology?

First, make sure you have well defined objectives. State the current benchmarks  of the related technology in your industry are and how you you would like to exceed them. On a side note, the objective can be to catch up to a competitor as long as their technology is proprietary.

Second, describe the initial research you did for available solutions. The CRA calls this due diligence (you can’t claim these hours). This includes internet and patent searches, talking to suppliers and other experts, or articles & manuals referenced. Make sure you explain why the available solutions are insufficient. For example, robotics could improve your process but are too expensive, therefore you sought to develop a cheaper solution in-house.

Third, document your findings. The CRA loves documentation and this is an area that is often overlooked. Make sure to save any emails, articles, search results, or other findings when you do your initial research. See my previous post The 5 Documents That Prove SR&ED for tips on documentation.


The remaining examples provided by the CRA focus on the methods applied to achieve your objective. Specifically, they want to know whether you follow the scientific method. The CRA defines this as “an approach that includes defining a problem, advancing a hypothesis towards resolving that problem, planning and testing the hypothesis by experiment or analysis, and developing logical conclusions based on the results.”

In a typical project this translates to selecting the key variables, evaluating alternatives for these variables, testing/prototyping the selected alternatives, and evaluating the results. this process is often repeated until the desired results are achieved or the project is abandoned.

Knowing, following, and documenting this framework makes for a great SR&ED project. Often companies with budgets and deadlines to meet don’t follow this method exactly and that is where a consultant, like me, is useful. Consultants help separate the business work from scientific development and put it into a concise report for the CRA.  Unfortunately, it is still up to you to prepare and save documentation produced during the project.


Focus on following and documenting these 2 aspects and the rest of the project will fall into place.

The 5 Documents That Prove SR&ED

Every SR&ED should be backed by documentation. The CRA often requests documentation to support a project. This can be financial and/or technical information. While financial information tends to be straightforward (financial statements, invoices, etc.), illustrating your technical process can be difficult. By matching documentation to the 5 steps of SR&ED you can ensure all the technical requirements are met.

Step 1a) Define Standard Practice

Before starting a project you probably search to see if there are any readily available solutions to your problem

Internet searches

The first place most people start is Google.  Articles showing that there is no standard practice or knowledge available that solves your problem.

To save the articles you can print to pdf or use a service like Evernote Webclipper. Once saved you can upload the evidence directly to RDBASE.


You may want to search patents to see if there is a product or process similar to what you are trying to create.  Google Patents is a great service that lets you search through millions of patents. Again, when you find something relevant, save & upload.

Inquiries to Experts

If you contact suppliers, customers, or other experts in the field save their response. Experts give a good sense of what knowledge is available to those in the field. Create a SR&ED folder in your email client and keep their responses.

Potential Components

Maybe you’ve found some parts or methods that you might be able to use but they don’t provide the full solution or are too expensive. Save the specs/article and make sure it illustrates why it’s not a full solution.

Competitive Products

Say your competitor has come up with a new product and you want to catch up to them. The solution isn’t readily available because it is proprietary to your competitor. Show what they’ve done and how you want to improve upon it.

Step 1b) Objectives > Standard Practice

Show how you plan to improve upon the current standard.

Look for a document that:

–          Outlines quantified objectives,

–          Discusses what the project is to achieve

–          Is dated near the start of the project

Possible sources of this documentation include:

–          Project planning documents

–          Emails

–          Meeting minutes

–          Design of experiments

Step 2 Uncertainty/Hypothesis

A test matrix or testing plan is a spreadsheet with variables/conditions to be tested as the column/row headers and the results of each test filling the spaces. Example:


Temperature 1

Temperature 2

Temperature 3

Time 1

Result 1-1

Result 1-2

Result 1-3

Time 2

Result 2-1

Result 2-2

Result 2-3

Time 3

Result 3-1

Result 3-2

Result 3-3

Showing how the variables were tested will prove:

– You were uncertain how to meet your objectives,

– had a systematic testing plan in place &

– you had a hypothesis how you could reach your goals.

This is the most important of all the supporting documentation. Not only with this backup all of your activities, it will serve as the starting point for the SR&ED project.

Step 3a) Systematic Investigation

The objective here to illustrate the difficulties you had during the project.

Pictures/video are great for showing issues that occurred on the shop floor. Add notes to the pictures to tie them back to the variables and comment on what went wrong.

Other documentation such as worker logs, time sheets, test data, or results are great to back up the work done and the costs claimed but don’t always prove SR&ED.

Step 3b) Technological Conclusions

This can be the hardest to find because analysis of the results is often in the managers head or discussed verbally. What is required is a document stating what was learned from the experiments in relation to the variables tested. Look for this in:

– in-term or  final reports,

– emails discussing results ,

– or meeting minutes. 


The best thing to do is establish a good documentation practice and follow it during the project. This includes:

–          Monthly SR&ED project meetings/reports

–          A physical SR&ED folder or online storage such as RDBASE

–          An understanding of which documents relate to SR&ED

What is a “Hypothesis” for SR&ED

Null hypothesis 1

The practice of science involves formulating and testing hypotheses, assertions that are capable of being proven false using a test of observed data. The null hypothesis typically corresponds to a general or default position. For example, the null hypothesis might be that there is no relationship between two measured phenomena or that a potential treatment has no effect. The term was originally coined by English geneticist and statistician Ronald Fisher in 1935. It is typically paired
with a second hypothesis, the alternative hypothesis, which asserts a particular relationship between the phenomena.


Hypothesis testing works by collecting data and measuring how likely the particular set of data is, assuming the null hypothesis is true. For instance, a certain drug may reduce the chance of
having a heart attack. Possible null hypotheses are

“this drug does not reduce the chances of having a heart attack” or

“this drug has no effect on the chances of having a heart attack”.

The test of the hypothesis consists of administering the drug to half of the people in a study group as a controlled experiment. If the data show a statistically significant change in the people receiving the drug, the null hypothesis is rejected.

Testing for differences

In scientific and medical research, null hypotheses play a major role in testing the significance of differences in treatment and control groups. The typical null hypothesis at the outset of the experiment is that no difference exists between the control and experimental groups (for the variable being compared). Other possibilities include:

  •  that values in samples from a given population can be modeled using a certain family of statistical distributions.
  • that the variability of data in different groups is the same, although they may be centered around different values.


Given the test scores of two random samples of men and women, does one group differ from the other? A possible null hypothesis is that the mean male score is the same as the mean female score:

H0: μ1 = μ2
H0 = the null hypothesis
μ1 = the mean of population 1, and
μ2 = the mean of population 2.

A stronger null hypothesis is that the two samples are drawn from the same population, such that the variance and shape of the distributions are also equal.

A one-tailed hypothesis is a hypothesis in which the value of a parameter is specified as being either:

  • above a certain value, or
  • below a certain value.

An example of a one-tailed null hypothesis would be that, in a medical context, an existing treatment, A, is no worse than a new treatment, B. The corresponding alternative hypothesis would be that B
is better than A. Here if the null hypothesis were accepted (i.e. there is no reason to reject the hypothesis that A is at least as good as B), the conclusion would be that treatment A should continue to be used. If the null hypothesis were rejected, the result would be
that treatment B would used in future, given that there is evidence that it is better than A. A hypothesis test would look for evidence that B is better than A, not for evidence that the outcomes of treatments A and B are different. Formulating the hypothesis as a “better than” comparison is said to give the hypothesis directionality.


Quite often statements of point null hypotheses appear not to have a “directionality”, namely, that values larger or smaller than a hypothesized value are conceptually identical. However, null hypotheses can and do have “direction”—in many instances statistical theory allows the formulation of the test procedure to be simplified, thus the test is equivalent to testing for an exact identity.
For instance, when formulating a one-tailed alternative hypothesis, application of Drug A will lead to increased growth in patients, then the true null hypothesis is the opposite of the alternative hypothesis, i.e. application of Drug A will not lead to increased growth in patients (a composite null hypothesis). The effective null hypothesis will be application of Drug A will have no effect on growth in patients (a point null hypothesis).

The Testing Process 2

In the statistical literature, statistical hypothesis testing plays a fundamental role. The usual line of reasoning is as follows:

  1. There is an initial research hypothesis of which the truth is unknown.
  2. The first step is to state the relevant null and alternative hypotheses. Specifically, the null hypothesis allows to attach an attribute: it should be chosen in such a way that it allows us to conclude whether the alternative hypothesis can either be accepted or stays undecided as it was before the test.
  3. The second step is to consider the statistical assumptions being made about the sample in doing the test; for example, assumptions about the statistical independence or about the form of the distributions of the observations.
  4. Decide which test is appropriate, and state the relevant test statistic T.
  5. Derive the distribution of the test statistic under the null hypothesis from the assumptions. In standard cases this will be a well-known result. For example the test statistic may follow a Student’s t distribution or a normal distribution.
  6. The distribution of the test statistic partitions the possible values of T into those for which the null hypothesis is rejected, the so called critical region, and those for which it is not.
  7. Compute from the observations the observed value Tobs of the test statistic T.
  8. Decide to either fail to reject the null hypothesis or reject it in favor of the alternative.

The decision rule is to reject the null hypothesis H0 if the observed value Tobs is in the critical region, and to accept or “fail to reject” the hypothesis otherwise. An alternative process is commonly used:

6.  Select a significance level (α), a probability threshold below which the null hypothesis will be rejected. Common values are 5% and 1%.

7. Compute from the observations the observed value tobs of the test statistic T.

8. From the statistic calculate a probability of the observation under the null hypothesis (the p-value).

9. Reject the null hypothesis or not. The decision rule is to reject the null hypothesis if and only if the p-value is less than the significance level (the selected probability) threshold.

Choice of testing process

The two processes are equivalent. The former process was advantageous in the past when only tables of test statistics at common probability thresholds were available. It allowed a decision to be made without the calculation of a probability. It was adequate for classwork and for operational use, but it was deficient for reporting results. The latter process relied on extensive tables or on computational support not always available. The explicit calculation of a probability is useful for reporting. The calculations are now trivially performed with appropriate software.

Common test statistics

In order to address the null hypotheses a series of analytical methods are applicable:

One-sample tests are appropriate when a sample is being compared to the population from a hypothesis. The population characteristics are known from theory or are calculated from the population.Two-sample tests are appropriate for comparing two samples, typically experimental and control samples from a scientifically controlled experiment.

Paired tests are appropriate for comparing two samples where it is impossible to control important variables. Rather than comparing two sets, members are paired between samples so the difference between the members becomes the sample. Typically the mean of the differences is then compared to zero.

Z-tests are appropriate for comparing means under stringent conditions regarding normality and a known standard deviation.

T-tests are appropriate for comparing means under relaxed
conditions (less is assumed).

Tests of proportions are analogous to tests of means (the 50% proportion).

Chi-squared tests use the same calculations and the same probability distribution for different applications:

  • Chi-squared tests for variance are used to determine whether a normal population has a specified variance. The null hypothesis is that it does.
  • Chi-squared tests of independence are used for deciding whether two variables are associated or are independent.
  • Chi-squared goodness of fit tests are used to determine the adequacy of curves fit to data. The null hypothesis is that the curve fit is adequate.

F-tests (analysis of variance, ANOVA) are commonly used when deciding whether groupings of data by category are meaningful. If the variance of test scores of the left-handed in a class is much smaller than the variance of the whole class, then it may be useful to study lefties as a group. The null hypothesis is that two variances are the same – so the proposed grouping is not meaningful.

Sample size

Statistical hypothesis testing involves performing the same experiment on multiple subjects. The number of subjects is known as the sample size. The properties of the procedure
depends on the sample size. Even if a null hypothesis does not hold for the population,
an insufficient sample size may prevent its rejection. If sample size is under a researcher’s control, a good choice depends on:

  • the statistical power of the test,
  • the effect size that the test must reveal and
  • the desired significance level.

The statistical power is the probability of rejecting the null hypothesis when it does not hold in the population (i.e., for a particular effect size).The significance level is the probability of rejecting the
null hypothesis when the null hypothesis holds in the population.

According to published theory, “Generally fewer than 30 trials puts any conclusion at risk.” 

Further issues in health science studies

Biostats uses basic statistics only as a foundation.

Biological variability results in developing stats  applications well beyond those that have been listed & generally requires advice from a biostats practitioner.

Each study has to tailor its stats tools to the overall objectives & intended approach of the study (e.g.,

  • different applications/premises used to identify
  • causal agents affecting health in epidemiology vs.
  • determining potential health outcomes in treatment studies, etc.).

then study specific objectives including,

  • calculation of adequate population size,
  • methodology (inclusion/exclusion criteria, type & number of biomarkers, cohort assignment, etc.) &
  • statistical analyses methods which are inextricably linked.

A study protocol that incorporates all these facets prior to embarking on data collection is a key component of an eligible study.

 The RDBASE SR&ED Consortium © 2012 Simplifying the SR&ED Process

Murray Arlin Dentistry PC – adequate documentation

Copies of the judgments are available from the Tax Court of Canada’s website.[1]

Murray Arlin Dentistry PC[2]


The appellant is a professional corporation that operates the dental practice which specializes in implants.

Fifteen years ago, Dr. Arlin purchased a computer software program called the Tritan Dental Implant Management System, which is designed to track the success rate of various types of dental implants.

Dr. Arlin uses the software to compare the success rate of implants in different circumstances. Some of the variables relate to the patients’ circumstances (e.g. smokers versus non-smokers) and other variables to the characteristics of the implant device.

The program contains approximately 200 potential inputs for every implant. According to the testimony, Dr. Arlin uses about 50 of these. Currently he has records for approximately 12,000 implants.

Dr. Arlin believes that by studying this data he can provide a useful addition to scientific knowledge.

Dr. Arlin estimated that he spent 350 hours per year on SR&ED since Fridays were spent on research when he does not see patients.

Evidence of experimentation or analysis

Dr. Arlin testified that he updated his research for all of his lectures.

The judge also noted that;

a) this testimony was very brief

b) should have provided greater detail and documentary support &

c) many of the lectures were:

  • not given to implant specialists &
  • had a marketing component.
  1. Whether there was systematic investigation &
  2. whether the allocation of Dr. Arlin’s time was reasonable.
Relevant legislation and analysis:

A significant focus at the hearing was on the requirement of “systematic investigation” in the definition of SR&ED[3] in Income Tax Act.

The CRA argued the research is not sufficiently documented to qualify as “systematic investigation” since;

a)       Dr. Arlin “failed to develop specific hypotheses prior to the data collection &

b)       there is insufficient evidence of time spent by Dr. Arlin on research in the relevant years.

Ruling & rationale: loss due to lack of documentation

The judge;

  1. was “reluctant to agree with” the requirement for “hypotheses [to be] determined prior to the data collection” however,
  2. “the main problem … very little detailed evidence regarding the analysis done in the years at issue and the time spent.”

She stated that,

“the Tritan program is designed to present comparative tables at the press of a button. The actual time spent on applied research potentially might be very small….

In order to support the appellant’s claims, the evidence as to actual research done, and the amount of time spent, would have to be much more detailed.”

Implications and author’s commentary

Though the judge did not require pre-stated hypotheses these might have helped the situation as far as relevant evidence.

The biggest disappointment in this case was the claimant’s inability to provide any real evidence of experimentation or analysis.

We are told they provided a single research article which was published in 2007 in order to support claims for the 2007 and 2008 taxation years.  Clearly the 2007 article could NOT have dealt with the 2008 work and perhaps not even 2007 work.

Results vs. Conclusions:

Basically Dr. Arlin’s system was able to illustrate “what” happened however he did not appear to have any written evidence attempting to document;

  • Why  these results occurred &
  • How any conclusions were formulated.

Evidence examples

The following list illustrates the types of evidence which are typically used to substantiate these types of claims. If Dr. Arlin had provided any of these they would have been excellent supporting documentation.

  • Notebooks – dated daily with brief, point form notes of hypotheses, related analysis & time spent
  • Emails – correspondence with the suppliers & colleagues regarding any  hypotheses & analysis.
  • Test Reports – any queries from the Tritan system which were used to analyze hypotheses.

Defining the SR&ED hypotheses

This is probably one of the most important and misunderstood sections of the SR&ED process.

To address this issue further we have outlined some of the key issues and opportunities in defining the “hypotheses for SR&ED purposes.”

[1] Tax Court of Canada website []

[2] Murray Arlin Dentistry Professional Corporation v. The Queen – Tax Court of Canada,  2012 TCC 133, Informal procedure

[3] Income Tax Act subsection 248(1)


The federal government’s 2012 budget included 4 minor changes to the SR&ED tax credit program:

SR&ED changes in March 29, 2012 Federal budget

Year change proposed to start (prorate)






Full Effect


Federal ITC rate (non-CCPC)





Subcontractor costs (% eligible)





Rate to calculate proxy (overhead)





Capital equipment (% eligible)