Fix Canada’s Primary R&D Program

Summary

Scientific Research and Experimental Development credits (SR&ED) are the single largest federal support for business R&D, allocating almost $5 billion a year¹. But, the program faces significant issues: 

• Program administration is exceptionally cumbersome. Applications must be done every year, after the research is finished, using long-form descriptions of the work completed. This creates significant risk and uncertainty leading companies to rely heavily on consultants to navigate the process. 
• Many of the biggest users of SR&ED are companies that do not need the money. These are either large corporations with steady income for research or foreign companies that do research in Canada to take advantage of the program but use the IP created to profit in other jurisdictions.
• The program is exclusively available for work within a narrow description of research and experimentation and there is no support for commercialization. In reality these lines are not clear cut and this contributes to the situation where most Canadian-owned IP ends up being used abroad^{2 3}.

By learning from countries with best in class research credit programs like the Netherlands and Norway we can solve these issues with three targeted policy changes:

1. Pre-approve Applications. Create a pre-approval step that can be filled out in an afternoon ensuring that - if the application is successful - the project is approved for up to 5 years, removing uncertainty.
2. Remove Wasteful Spending. Reduce the SR&ED refundable rate offered to large companies and stop offering SR&ED credits to foreign companies unless the IP stays in Canada.
3. Support Commercialization. Make it possible to claim some commercialization costs for a project as part of the SR&ED program.

Background and Motivation

Today to receive SR&ED credits a company submits a claim following the tax year using form T661. This form requires lengthy narratives describing the company’s R&D to show their “scientific uncertainty,” “systematic investigation,” and “technological advancement.” 

Alongside the description companies submit expenses related to the R&D explicitly excluding areas of essential work like “routine data collection”, “market research” and “the ordinary testing or inspection of materials, devices, products or processes.” 

Following submission the CRA may audit or review. Estimates put combined review/audit rates at 25–40%⁴ with a reviewed application often stretching into many months. This system creates significant problems.

Firstly, sending in applications is time-intensive and risky. The timeline means that between the point that a company conducts R&D and the point it has confidence it will receive a credit can be the better part of a year or more. This discourages starting any project that has a chance the SR&ED credit will not be approved. The standards for a claim being accepted are also unclear meaning that - especially for first time filers - there is significant risk to applying for every new project. And refiling every year creates a large paperwork burden. 

As a result over 80% of claimants use consultants⁵. This wastes a huge portion of the funding that should be going to innovation. Many work on contingency fees that amount to a third or more of the credit issued. This setup has the worst impact on small claims where consultants can turn a $30,000 SR&ED claim into $20,000 for the company. The result is that billions of dollars in taxpayer funding meant to support R&D is being diverted into consultant revenues.

Secondly, much of SR&ED credits go to support companies that don’t need it. Some of the biggest recipients of SR&ED are large companies that already have substantial cash flow⁶ so they rely less on SR&ED to fund incremental R&D. And many international firms use SR&ED to fund the creation of IP that is ultimately owned by a foreign entity essentially expropriating the IP to create benefits in other markets.

Lastly, the ultimate aim of SR&ED is to spur economic growth by incentivizing companies to invest in productivity raising activities. But, deciding ahead of time what counts is exceptionally difficult. For example, since SR&ED was developed it has become clear that some activities the program explicitly rules out like continuous improvement, leveraging new data, preparing patents and even support functions like product discovery are essential parts of innovation. 

Real World Solutions

In European nations and around the world these issues have already been addressed with pre-approvals, requirements for IP ownership by domestic companies, and expanded credit eligibility.

Simple online portals and pre-approvals are standard best practice. The New Zealand Research & Development Tax Incentive offers “General Approvals” covering up to three years of R&D⁷; the WBSO program in the Netherlands uses a digital portal to pre-approve projects with nearly 90% of applications processed in three months⁸; and, in Spain for Informe Motivado Vinculante (IMV) the Ministry of Science issues binding rulings on eligibility, which the tax authority must respect⁹ so companies know they will not face disputes. 

Preventing foreign ownership of resulting IP is a recognised issue. In Belgium the amount of credits offered under the Innovation Income Deduction (IID) depends on “nexus” criteria that require R&D and resulting new IP stays in the country.

To solve problems around commercialization many European countries and the UK already have expanded criteria. In France the government created the Crédit d’Impôt Innovation, CII alongside their traditional R&D support with Crédit d’Impôt Recherche, CIR¹⁰. This allows SMEs to claim credits not only for scientific R&D but also for prototype development, pilot testing, and design activities directly tied to commercialization. 

What Must Be Done

We can reform the SR&ED process to deal with its largest problems with three simple changes while keeping the majority of the system intact:

1. Create an online portal to allow engineers and scientists to apply for SR&ED using simple language they are familiar with and get projects pre-approved for up to 5 years.
2. Remove the ability for foreign companies to use SR&ED credits to pay for IP that is then owned and benefitted from overseas.
3. Expand the ability to use the credits on essential R&D work like prototype refinement and getting regulatory approval.

Adopt a Pre-Approval Flow to empower engineers and scientists to apply for SR&ED directly. Under this model, a company would submit simple hypothesis-driven templates in advance, choose multi-year approval periods, and receive binding approval certificates. Annual filing would be reduced to reporting costs and checking a “pre-approval” box on Form T661 that could be handled by normal accounting procedures. Here’s how it would work:

1. When starting a new R&D project, a company would go to their CRA Business Account to fill out a pre-approval form. This form would be a simple hypothesis-driven template where the company lists: uncertainty, hypothesis, method, expected outcome. Scientists and engineers could complete these directly using language and descriptions they are already familiar with from their training.
   
   
2. As part of the application the company would select a pre-approval period of 1–5 years based on how long they think the research will likely continue for.
   
   
3. Following the submission the CRA would review the application and issue a binding Pre-Approval Certificate covering the full period. This would include a Certificate ID that can be used to identify the project as part of the company’s annual filing.
   
   
4. When it comes time to file the company will still submit Form T661, but would include their Certificate ID and check a “Pre-Approval” box, alongside reporting total eligible expenditures. No technical narratives are required. If a company is pre-approved the CRA cannot re-litigate technical eligibility, though it may still audit costs.
   

The Income Tax Act could be amended to authorize the CRA to issue the Pre-Approval Certificates with the determination being binding for the purposes of Sections 37 and 127. While associated regulations could be updated to describe the pre-approval process. It could then be implemented by extending the CRA’s My Business Account with a dedicated SR&ED Pre-Approval portal. Providing a set of standardized templates with examples to encourage brevity and discourage unnecessary attachments. Finally, updating form T661 to add fields for a Certificate ID and Pre-Approval checkbox.

Change the rules for which companies can claim SR&ED. To make sure that SR&ED credits are being used where they are most impactful and that the resulting IP stays in Canada make two changes to applications. First, progressively reduce the refundable portion for the largest firms, e.g. above $250m in revenue. This avoids creating a scaling bottleneck while improving efficiency of funding.

Second, amend the rules to make it clear that the entity performing the R&D must also own the underlying IP. Under this structure if foreign companies claim credits for work done in a Canadian subsidiary, that subsidiary (or a related Canadian subsidiary) must own the resultant IP. This would mean that the company which benefits from the wealth creation of new IP remains in Canada and ensures the value created from that IP would eventually be taxed in Canada as a royalty, license or capital gain. 

Update form T661 with a new “commercialization expenditure” section. When companies apply for SR&ED, as long as the work is tied to the experimental project, they will be able to claim expenses here related to other forms of work like “prototype refinement for production”, “regulatory approval costs”, “pilot projects with early customers”, “IP registration/protection”, and “specialized staff for scaling the initial results into a product”.

Common Questions

Won't this make it easier for companies to abuse the system? Pre-approval actually reduces abuse. Companies must describe their research upfront when they can't know the results. This stops after-the-fact storytelling that happens today. The CRA gets better oversight of what research actually happens.

Will consultants lose their jobs under this system? Some SR&ED consultants will shift to helping with the pre-approval process and other tax services. But taxpayers shouldn't subsidize an industry that exists only because the government program is poorly designed. The goal is supporting research, not consultant revenues.

Doesn't limiting foreign companies hurt Canada's competitiveness? Canada will still attract foreign investment. But taxpayers should get value for their money. If a foreign company wants SR&ED credits, they must create IP that benefits Canada. This is already standard practice in Belgium and other competitive countries with high R&D spend.

Conclusion

SR&ED credits are Canada's biggest R&D program, but they've become a complex process that props up a huge consulting industry, rewards the wrong companies, and ignores commercialization. Three simple changes can fix this: pre-approve projects to cut consultants, focus credits on companies that keep IP in Canada, and support the full innovation process from research to market. These reforms will help turn SR&ED into what it was meant to be: a driver of Canadian innovation and economic growth.

‍

If these ideas resonate and you want help to build Canada sign up to join the movement today to be the first to hear about events and volunteer opportunities.