An Open Response to the Second Round of SR&ED Consultation
An Open Response to the Second Round of SR&ED Consultation
Specific recommendations for improvements to the SR&ED program focused on incentivizing domestic commercialization of Canadian IP
Context
Early this year, the government began a consultation process with the goal of overhauling the SR&ED program, seeking to effect cost-neutral restructuring to help address Canada’s deficit of patents and IP-related payments and a host of other complaints with both the administration and impact of the program. The innovation community responded in lockstep, with recommendations from independent corners of the innovation community aligning strongly on all major points (see my response, that of CCI, and those of Quantum Industry Canada, for example).
Recently, a second call for consultation was released, focusing on a set of specific questions relating to the feedback provided by the innovation community in the first round. The content of these questions is encouraging in that the subject matter indicates that the key points have been heard and the proposed solutions are at least being explored. Below, I have written an open letter that responds to the latest round of questions. Whether or not you agree with my assessment, I strongly encourage you to respond to the consultation call. This is a once-in-a-lifetime opportunity to enact substantial changes to the single most important innovation support program in Canada.
Let’s not waste it.
An open response to “Cost-neutral ways to modernize and improve the SR&ED program”
To generate the email with a filled in header in your default mail client, click here
To whom it may concern,
I am a Canadian biophysicist and entrepreneur operating at the interface between academic research and early-stage commercialization of innovative technologies arising from Canadian academic R&D. In the course of my career I’ve been involved in the innovation process in many ways. As an inventor and researcher, I am responsible for generation of 27 patents across 4 patent families, filed globally. As founder of a nanotechnology company commercializing part of this portfolio, I used or interacted with every Canadian innovation support program for early-stage commercialization, including SR&ED. In my role as Entrepreneur-in-Residence in the Faculty of Science at the University of Ottawa, I serve as a mentor and advisor to other scientists and engineers in the early stages of commercializing their inventions. Finally, as a policy critic, I attempt to turn these experiences into productive inputs to the innovation policy discourse.
The lens through which I view the challenges faced by Canadian innovators is focused on the early stages of technology commercialization, a stage where companies are often IP-rich but cash-poor and are acutely vulnerable to the pressures that lead valuable Canadian IP portfolios to exit the country before realizing their economic potential. In other words, I focus on the stage of development at which the foundation is laid for Canada’s economic future, and for which SR&ED is of existential importance. It is my belief, backed up by first-hand experience from multiple perspectives, that this is the stage of building during which the majority of Canadian IP is lost, and that this is the stage of building during which resources such as SR&ED can achieve the largest long-term return on investment, if delivered effectively.
I previously wrote a response to the first round of SR&ED consultation, the full text of which can be found here. Below, I provide specific answers to the questions asked in the second round of consultations. These responses should be considered in the context of the specific recommendations made in the first round.
1. What are some of the challenges faced by research-and-development-performing small- and medium-sized Canadian public corporations when it comes to financing?
Small companies, especially those operating pre-revenue, face major challenges with cashflow. Securing funding in Canada, from any source, is particularly challenging for pre-revenue companies that are conducting R&D, which is to say almost all Canadian companies attempting to commercialize academic research. Most government support programs that support R&D costs, including NRC IRAP, Mitacs, NSERC Alliance, and others, all require either revenue minima for eligibility or fronting of significant cash resources, requirements that exclude IP-rich, cash-poor companies in the early stages of development. Given the risk-aversion of Canadian venture capital relative to the United States, this often forces companies to leave the country to develop elsewhere rather than attempting to find support domestically.
Among Canadian innovation support programs, SR&ED is very nearly unique in that it does not require revenue minima to access. However, the fact that cash is only received after expenditure and the subjectivity involved in application review by non-expert CRA personnel imposes a soft version of the same requirement: companies must front significant cash, without a guarantee of getting SR&ED credits back. All of this combines to make the program less effective than it could be, and disincentivizes calculated risk-taking on valuable, tax-payer funded IP portfolios. While SR&ED financing is available, the high interest rates associated with this route severely dilute the impact of the credit if used by as much as 30% (see my previous discussion of this in point 3).
This can be remedied quite simply, by moving to a pre-approval process with spot-checks to ensure compliance. By providing cash on an ongoing basis, SR&ED can provide critical cashflow and flexibility to innovative Canadian companies in their most vulnerable stage, laying a foundation from which the next generation of companies can be built to support long-term Canadian prosperity.
2. To avoid any potential disincentives to growth, would entrepreneurs favour a program with one single rate accessible to all, even it if means somewhat lower support for small Canadian-controlled private corporations?
I am generally of the opinion that support for industry should be proportional to the need, which is to say that it is appropriate to provide more support for a very large number of companies earlier in their development rather than to support large incumbent firms.
The main criticism of the reduced rate above $10M return is that it disincentivizes investment in R&D by larger firms once they pass this threshold. With a well-designed overhaul of SR&ED, this need not be the case. If the program provides a competitive advantage to small companies over large ones, then larger incumbent companies will face more competition and will be motivated to innovate to stay relevant. The only reason that large companies do not feel pressure to innovate in Canada currently is because Canada provides no support for emergent challengers, and so there is no need to innovate to stay competitive. If this changes, so too do the incentives for larger firms, irrespective of the level of support provided. In short, I do not see any need to change the progressive structure of the SR&ED program; only how it is delivered.
A good example of the kind of ecosystem that can be built through concentration of support on the early stages of building in the Israeli Innovation Authority, which provides extensive support for the early stages of building while still have a strong presence of foreign direct investment by large multinational anchor firms. A more detailed analysis of this model can be found here.
3. How should the concept of "Canadian" public corporations be defined, should the government proceed with measures to improve access to the SR&ED program's enhanced credit for Canadian public corporations?
A corporation should qualify as Canadian for the purposes of SR&ED eligibility if and only if it is headquartered in Canada and is controlled by Canadian individuals (in the sense of a majority of shareholder voting power or by a majority Canadian presence on the board of directors, as applicable). In the case of a subsidiary, such a subsidiary should be considered Canadian if and only if the controlling parent corporation is Canadian.
While there is no doubt nuance to consider in the above, the intention of the definition is to ensure that branch-plants of foreign multinationals cannot benefit from SR&ED, or at best do so at a reduced rate. Enough subsidies exist in Canada for such entities without redirecting SR&ED credits away from innovation Canadian companies that actually need them to get off the ground.
It is clear from the numbers that such a step is necessary. Various reports have found that a majority of the SR&ED credit goes to a small number of firms, many of which are Canadian controlled only in name. This is of limited value to Canada in the long term, as it is clear from Canadian economic stagnation that the strategy of branch plant economics is ineffective in promoting long-term economic prosperity.
4. The SR&ED program currently has rules to prevent the multiplication of the expenditure limit by Canadian-controlled private corporations with common control. If enhanced support were extended to public corporations, how should relationships among legal entities be delineated?
In general, a company large enough to spin out multiple entities for the sake of taking advantage of additional tax credits does not need the extra support. Eligibility and expenditure limits should apply to the controlling parent company, rather than to its subsidiaries separately, to avoid multiplication of the limit. To qualify independently for its own SR&ED limit, an entity should be independent. The response to question 3 above provides additional rationale.
5. Current global initiatives rely on accounting concepts of relationship and control to determine whether entities are included in a large business corporate group. Should existing international practices of this sort be adapted for determining relations for public corporations in the context of the SR&ED program?
In general, and considering the answers to the previous two questions, such practices will be a necessity. By ensuring that parent corporations and large business corporate groups are subject to a single shared limit on SR&ED credits, the credit will be much more widely distributed to a larger number of smaller independent firms, laying the foundation for a future version of Canada’s economy built on Canadian innovations and technology rather than branch plants of multinationals that result in effective if not nominal export of Canadian IP.
6. What is the optimal size-based metric (e.g., taxable capital employed in Canada, revenue) to phase out enhanced support for public corporations, including those in a corporate group?
If the goal of SR&ED is to encourage investment in development and ultimately commercialization of Canadian R&D, then it is critical that the credit be focused primarily on small companies. The reasoning for this is simple. A small company is only going to hold IP that they are actively using and which they reasonably believe will generate revenues, because filing and maintenance of an IP portfolio is a costly endeavour. A large firm, by contrast, often patents anything and everything for the sake of developing a large moat, with no intention to use it as anything other than a lawsuit deterrent. In other words, IP held by small companies is more likely to result in long-term economic benefit than IP held by large ones.
Taxable capital employed is problematic as a measure of size. A deep tech company that (somehow) finds significant VC support in Canada could employ substantial funding while still being fully focused on value creation through intellectual property (see for example Canada’s plethora of quantum companies). Such companies should not be excluded from the credit - on the contrary, these are the companies that are most important to support.
Company size as measured by employee count is a reasonable, if imperfect metric by which to gauge SR&ED eligibility, provided that it considers the size of the group of companies as defined in the responses to questions 3, 4, and 5, and not to the size of subsidiaries individually. Where company size breaks down is in cases where large companies are formed on the basis of investment rather than revenues - a rarity in Canada, but something that could become more common if the innovation support reforms are executed well.
Gross revenue is to my mind the cleanest of the options presented. A company that has revenues has at least passed the point of pure value creation through intellectual property which is the stage at which SR&ED should be concentrated in the first place, and has achieved economic impact with that IP. Using revenue as a means to taper off SR&ED support is reasonable. This would also promote the formation of anchor firms in the long run: because companies that were previously innovating and generating IP using SR&ED will no longer be able to access the program, they will favour making strategic minority investments in independent companies as a means of IP creation over doing so in house. The Israeli Innovation Authority provides a good example of successful execution of a similar inventive structure.
The intention is that support should be maintained at the highest level until a company has achieved revenue generation from a portfolio of intellectual property. To assess the specific revenue breakpoints to use, simple gross revenue cutoffs should be developed using data available to the CRA and in consultation with Canadian SMEs that have developed or are developing intellectual property with this as the guiding principle. I would suggest basing this primarily on the one truly innovative industry in which Canada leads: quantum computing and quantum-related technologies. Canada has a concentration of these, they are representative of deep tech generally, and have clear long-term value despite for the most part being pre-revenue (or at least, pre full potential revenue).
7. How does refundability under the SR&ED program influence investment decisions and planning? To what degree would Canada become a more competitive location to undertake research and development (R&D), compared to other jurisdictions, if credits earned at the general rate were partially or fully refundable?
The refundability of SR&ED is critical to its value. As noted in the answer to the first question, cashflow is the biggest challenge to intellectual property development in the early stages. Not only should the SR&ED credit be fully refundable to ensure Canadian competitiveness relative to other jurisdictions as a place to conduct innovative R&D, it would be of further value if the refund was provided incrementally throughout the year rather than as a single lump sum after the fact.
In my previous recommendations I suggest integrating SR&ED with the payroll tax submission system so that it is calculated and remitted monthly instead of annually, as part of an administrative process that already happens monthly. If combined with pre-approval or default approval of SR&ED projects, this would dramatically enhance the value of the credit, both directly, by providing critical cashflow, and indirectly, by obviating the need for costly SR&ED consultants and financing, while at the same time reducing administrative overheads and uncertainty.
8. Would it be preferable that the government make the general rate refundable, but at a reduced rate? What would be an acceptable trade-off in this regard?
Given the current inefficiency in the SR&ED program, in which as much as 30% of paid credits are redirected to SR&ED lenders and/or consultants and a majority of the credit is paid to large multinationals that do not need it, there is plenty of room for the refund rate to come down slightly while still delivering greater value overall.
As long as rate reductions are combined with reforms that eliminate these inefficiencies in such a way that the overall impact is net-neutral, then there would be no harm in an overall rate reduction. However, rate reductions without addressing existing inefficiencies through structural reform would be harmful and counterproductive.
9. In your view, should SR&ED-eligible activity be broadened from the existing OECD definition of SR&ED, generally used by Canada and other countries offering R&D tax credits? If so, how would you propose to amend the current definition? Why would any additional activities warrant government support?
I very strongly believe that SR&ED-eligible activities should be broadened to include activities relating to all of the following:
Commercialization activities, including hardware and software product development and engineering based on SR&ED-funded IP
Curation and collection of datasets for AI/ML training
Intellectual property filing and protection, including (especially) patenting in foreign jurisdictions
The rationale for 1 is simple: development of an intellect property portfolio should not be the end goal. By ending support at the development phase, SR&ED actually promotes export of the resulting portfolios and/or use of IP portfolios as an unproductive lawsuit deterrent (see the response to question 6), resulting in Canadian taxpayers paying for development of IP in Canada that does not get commercialized here. Once development of IP is complete, there remains significant work in order to exploit it commercially. IP must be protected, probably broadly given that Canadian companies generally need to export to be relevant, and products must be developed before revenue can be generated.
The rationale for point 2 is developed in the answer to question 11, while the rationale for point 3 is elaborated in the answer to question 12.
While these additions no doubt will greatly increase the scope of eligible activities, cost neutrality can be maintained through elimination of support for the largest claimants per the responses to questions 4, 5, and 6. SR&ED can and should support a very larger number of primarily small firms to develop IP from idea to product export, rather than from idea to patent export and/or shelving as part of an IP moat as is currently the case.
10. Can you provide specific examples of activity that you think should be eligible for the SR&ED program that are not currently eligible? Would such a change bring additional predictability to claimants?
As noted in the previous question, eligible activities should be expanded to cover the following activities:
Commercialization activities, including hardware and software product development and engineering based on SR&ED-funded IP
Curation and collection of datasets for AI/ML training
Intellectual property filing and protection, including (especially) patenting in foreign jurisdictions
By including these activities in SR&ED-eligible expenses, claimants will have the additional certainty that developing innovative intellectual property will be a worthwhile exercise, since they can depend on having support all the way through the process to economic impact instead of being left with an unsupported patent portfolio or set of ideas without resources to protect them, as is currently the case when SR&ED support ends.
11. How could the SR&ED program be enhanced to support businesses conducting R&D in the digital age, particularly in respect of software development and the emergence of artificial intelligence?
As noted in the list of suggested supported activities in the responses to questions 9 and 10, by far the most time-consuming (and valuable) element of AI development is assembly of datasets to support model training. Control of high-quality, human-generated, curated data is the only defensible moat in the AI space, as the models themselves are relatively easy to reproduce. This is evidenced by, for example, the willingness of Meta and others to release their model weights in an open format: they control the training data, and so any competitor that does not have an equivalent dataset already cannot hope to do any better. The firm that controls the data required to train the largest foundation model will control and dictate the direction and pace of AI development in their respective application space.
While SR&ED support for software development should be made much less subjective generally, data collection and curation of datasets useful for model training is the key element of IP generation to support to ensure Canada has a presence in the ongoing and accelerating development of AI.
12. To what extent do businesses face financial challenges and trade-offs in protecting their intellectual property (IP) in Canada and abroad? Would it be appropriate for the government to provide additional support to these activities under the SR&ED program? If so, what would be a cost-effective approach?
As noted in the list of suggested supported activities in the responses to questions 9 and 10, protection of intellectual property is an onerous cost of doing business and a significant barrier to entry that disproportionately limits small companies as compared to larger ones. Because Canadian firms often must be exporters from the start in order to compete, patent protection is required in many jurisdictions, the cost of which can easily balloon beyond the means of small companies. This limits the ability of Canadian firms to compete globally, which in turn reduces Canada’s attractiveness as a place in which to establish a business.
Only very recently have government support programs for intellectual property arisen, and all have serious issues that prevent them from realizing the desired national impact: IRAP IP Assist does not pay for patent filing, CanExport SMEs requires revenue minima to access, and IPON, while they do excellent work in the early stages, are limited to one province in their potential impact.
It would be highly valuable for Canadian SMEs if SR&ED covered part of the cost of patenting intellectual property developed in the course of SR&ED-funded R&D. Reimbursing for patent filing costs at the prevailing SR&ED rate is a simple change that would greatly enhance the ability of Canadian SMEs to compete globally.
Of course, it is reasonable to build in a mechanism by which to ensure that Canadians benefit from the use of their tax dollars in the long term. For this purpose, an approach similar in principle, if not in implementation, to the Bayh-Dole Act in the US could be considered, administered by the Innovation Asset Collective, providing the Canadian government or some part thereof a means by which to access IP supported through the SR&ED program and use of public funds. Such a requirement would also disincentivize claiming of such support by companies that do not actually need it and/or who intend to export it or use it only as part of a moat.