Thoughts on "Strengthening the Impact of Foreign Investment on Domestic Innovation"

An excellent report by ICTC makes a strong case for a careful balancing of the value of IP retention against foreign direct investment

In this newsletter I write a lot about the importance of retaining IP developed through academic research as the basis for long-term economic growth via deep tech commercialization. Given this, one could be forgiven for assuming that I am generally against foreign direct investment (FDI) as a means to build the Canadian economy. Not so. While I am generally of the opinion that Canada’s approach to FDI is problematic currently, FDI has a very important place in the development of any economy, especially a small, open one like ours. However, the details of how, when, and at what stage of development FDI is used make the difference between ensuring long-term value for Canada versus being effectively an IP republic.

I admittedly do not know the “right” answers to those questions, answers which are certainly moving targets. However, a 2022 report by Matthews and Rice via ICTC provides an excellent guide to how to approach thinking about these issues.

In addition to asking very relevant questions, the report reframes traditional thinking around a few key issues, specifically by making clear that challenges associated with market size, talent retention, and IP are intricately intertwined and fundamentally inseparable. The fact that Canada’s market is small and lacks local talent in many areas makes scaling a business difficult, which in turn incentivizes early acquisition and/or moving to more established ecosystems, exacerbating the talent problem. At the same time, IP is not limited to just patents—the contents of the brains of those who leave can and should be considered valuable IP (or at the very least, potential IP). Incentivizing retention of businesses, talent, and IP are therefore not three separate issues, but must be considered as a gestalt and addressed from all stages of the economic development and industrial policy pipeline if measures are to be effective.

Given the length of this article, it will need to be covered in parts. In this one, I focus on the six questions asked in the Executive Summary section on government financing, drawing on the content of the report and adding my own thoughts. This section begins around page 80. Future articles may examine more closely other sections of the report.

For those interested, I encourage you to study this report in some detail, as it contains a wealth of valuable insight backed by high quality data throughout, to which justice cannot possibly be done in a few thousand words. Even better, it goes beyond simply pointing out problems and suggests actionable solutions that are viable, strongly supported by a combination of forms of evidence, and provide a framework for continuing to think about the time-evolution of the problem.

FDI Attraction via Research Funding

“Canadian R&D funding is frequently used by FDI attraction agencies to attract greenfield investment in Canada. University IP commercialization offices, accelerators, university departments, and not-for-profit institutes also play a role in FDI attraction by forging R&D partnerships or licensing agreements. At least two distinct perspectives on this type of partnership exist: one, that FDI partnerships for R&D produce and commercialize IP in Canada; the other, that such partnerships result in an exodus of Canadian IP without adequate commercialization opportunities at home.”

Canadian research funding agencies love public-private partnerships as the basis for allocating research funding. The vast majority of my own research work, for example, was funded primarily through a series of partnerships with foreign multinational entities (FMNEs) over the past decade or so. In that time, I have seen and been involved in every possible permutation of commercialization activity arising from publicly-funded research, ranging from shelving of IP, to developing IP in Canada via the Canadian branch of an FMNE, to spinning out my own company to take IP out of the lab. Obviously, these provide varying degrees of value to Canada in the long run, and via very different mechanisms. Striking the right balance between these outcomes is a critical part of optimizing use of FDI in driving technological innovation, but Canadian academic research funding agencies are almost entirely hands-off with respect to use of IP and make little attempt to do so, as noted in the report:

“federally funded academic research programs, such as NSERC and the Canada Research Chairs program, default to partner institutions’ resident IP policies.”

In any R&D arrangement, who retains the rights to the IP that is created?

Canada is a very attractive place for research conducted by local branch plants of FMNEs, though a large variety of mechanisms of which I mention the largest here.

SR&ED is accessible to Canadian Controlled Private Corporations (CCPCs) that conduct R&D, which can include Canadian branch plants of FNMEs. Per 2017 data, as much as 23% (about $1B) of SR&ED was paid out to such entities. SR&ED supports research done by a private sector company, and as such any resulting IP is owned by the entity itself. While some metrics relating to IP filing are considered, no long-term outcomes of the results are effectively tracked.

NSERC, the tri-council agency responsible for most Canadian STEM research funding in post-secondary institutions, runs Alliance grants, a program that pairs an established industry partner and provides a 2:1 match on cash provided by the industry partner to a university partner for research (N.B. newly formed startups are not eligible by default as cost-sharing partners, but can be made so through some additional hoop-jumping). Alliance grants have been growing steadily over the past few years and paid out about half a billion dollars in 2023. Similar to SR&ED, CCPCs are eligible as cost-sharing partners independent of their parent company headquarters location. Due to the matching mechanism, there is a structural incentive here to favor large companies as Alliance partners given their ability to fund joint research.

At the same time, the vast majority (85% or more) of such partnerships give control of the resulting IP to the industry partner, either via direct assignment, a license, or the option to license (see tables 11-1 and 11-2 here).

This favoring of established businesses extends beyond directly funded research, with the report noting that

“One IP commercialization office noted that most researchers opted to license their IP to a pre-established company rather than create a startup due to risk aversion and lack of experience, and that the company they licensed to was often a FMNE that retained the IP in the end.”

It goes on to quote one of their interviewees who makes the key point succinctly:

“Universities are strapped for funding. A company comes in and offers lab equipment and training, and it has generally not seemed like a bad deal to sign over IP in exchange for all those benefits. Universities and researchers are not interested in commercializing and developing businesses out of the IP. But there’s a real trade off from the economic perspective of business in this country”

In sum, most Canadian research-based FDI incentives give control of IP directly to the partner, with little to no tracking of long-term outcomes, favoring immediate metrics like jobs created while the partnership lasts.

When, where, and how is that IP commercialized?

The report makes the point very clearly that while benefit to Canada associated with having FMNEs establish local branch plants is typically in jobs created, this is not the only (or, I would argue, even the most) important consideration. Tax benefits and income arising from the IP itself in the long run is another key economic impact consideration. As the report states:

“it is important to carefully weigh the relative value of potential jobs—including their potential spillover effects, if Canadians are trained well and then able to stay in Canada—versus potential revenue from IP in any transfer of IP ownership or R&D agreement.”

Without a clear value assessment of the long-term potential value of the IP, it is not easy to assess whether this tradeoff is worthwhile for Canada. IP sold at the research stage has far less value than IP that has been developed to the point that it has a proven market. At the same time, because of a variety of factors including but not limited lack of access to scaling capital and difficulty finding support in the early stages of building, Canadian startups are often acquisition driven, and typically exit (either physically or via acquisition) sooner than their American counterparts. Even if IP must be sold at some point due to difficulties scaling in Canada, it may still be beneficial to develop it further before doing so (the example of Israel provides a point of comparison that favors this view).

While this report does not offer a specific answer to the question of what path of IP commercialization is best from an economic development perspective, they do suggest a clear and nuanced question to ask that should inform further tracking efforts as policy evolves:

“For the purposes of this study, it would be more fruitful to assess the GDP impact of IP created in Canada but commercialized by a foreign entity with operations in Canada, versus IP held by CCPCs.”

(N.B. one must also be careful of the definition of CCPC, since branch plants of FMNEs can be CCPCs).

This is not at easy prospect, given that it requires collecting metrics from FMNEs on the long-term results of their licensing. This in turn requires a contractual obligation to do so, which in many cases must be imposed at the level of university tech transfer in the context of public-private partnerships, and strongly supports the inclusion of reporting on commercialization outcomes as a federally mandated condition of funding, a topic about which I have written before.

How prepared are Canadian researchers and startups to meaningfully commercialize their IP?

My last article covered this issue in more detail: Canadian researchers have historically not been well trained in management and commercialization of IP. While several national programs are beginning to address this gap, these are relatively new initiatives, the results of which will take time to be felt economically.

The problem of researchers ill-equipped to commercialize their IP is compounded and reinforced by the structural incentives previously noted for universities to favor established businesses as IP receptors.

The report presents a solution, as well: provincial agencies that provide the required IP literacy. Top-down approaches like ElevateIP program and the various provincial subsidiaries, including IPON, are beginning to fill this role. As previously noted, individual institutions are also beginning to address this from the bottom up, giving hope that Canada is on track to overcome this hurdle to research commercialization.

It is only a relatively recent thing that Canadian universities are even trying to favor startup companies as IP receptors, with most associated support structures being only a few years old. Assessing the effectiveness of these initiatives and course-adjusting in response to real-world feedback will require that we stay the course over several years, and track metrics that accurately correlate to target economic outcomes (things like 2, 5, and 10-year survival rates and revenues, as well as detailed information on access to IP).

Without support from a foreign multinational enterprise (FMNE), what parties would purchase or scale Canadian IP?

“One interviewee coming from the perspective of an accelerator felt that R&D partnerships with foreign FMNEs were essential to move Canadian solutions toward market readiness and expanded market opportunities. Nearly all interviewees conceded that in the absence of FMNE partnerships, there are not enough large investors or first buyers in Canada to scale most new technology companies. […] the “problem” with foreign acquisitions may be ‘less about foreigners absconding with Canadian IP than with domestic difficulties bringing innovative ideas to market.’”

I think it is probably true historically that most Canadian research commercialization happens via FMNEs, but I think this is something of a self-fulfilling prophecy: as long as we structurally favor these entities as IP receptors and provide no incentive or mandate to do otherwise, this outcome is not surprising. In other words, Canada is effectively stuck in a vicious cycle with respect to IP commercialization and IP retention, that will not be broken without direct intervention from the funding agencies.

As the report notes, this is not a particularly novel idea. Fraunhofer in Germany, Bayh-Dole in the US, and the Israeli Innovation Authority on which the now-probably-defunct CIC was based are all examples of this approach that have been to varying degrees effective in getting academic technology to market via domestic SMEs.

The report asks the key question:

“why should Canadian companies or FMNEs choose to remain in Canada if growing their company will be an uphill battle?”

This one is simple to answer: without structural incentives to do so, they shouldn’t.

What goals and side effects of FDI/university partnerships are there to consider other than IP?

Something that I do not often touch on is that there are worthy goals other than IP commercialization. As the report notes explicitly, strong incentives for FDI can have knock-on effects with respect to international relations and may present opportunities for benefits to Canada that are unrelated to the creation and scaling of technology companies, including provisioning of new buildings and research complexes for universities. These considerations may well outweigh the value of the IP that is traded in the process.

I would argue, however, that while that is true, it is not often the case that universities engaging in public-private partner-funded research are considering the broader political implications, nor are tri-council funding programs being managed with these knock-on effects as a conscious goal. If allowing IP to leave the country at the point of creation were an intentional part of a broader strategy for long-term economic growth I could probably get behind it, but I have seen no reason to believe that this is the case, and a decade of lagging innovation metrics supports this view.

What countries are we competing with to attract FDI, and how do our incentive structures compare?

I have written in the past about one of the suggestions that the report brings up: tying government grants for R&D to retention of the IP, on pain of paying back the grant, with interest. This is already done in many funding contracts, including (to varying degrees) SR&ED, NRC IRAP, SIF, and CanExport, among others. However, this approach raises the question:

“is it contradictory to support Canadian IP development, retention, and commercialization while keeping Canada an attractive place to do business for international players?”

The example of Israel proves that this needle can be threaded. In Israel, government grants are heavily used in the early stages of deep tech scaling, but many are paid back many-fold once companies are making enough money to enable them to do so. At the same time, Israel benefits enormously from FDI, with a huge number of FMNEs actively involved in supporting research commercialization. While Israel still has issues of IP exiting the country in the long run, it is typically developed to a much greater degree before it is acquired, accruing far more value for the country than a piece of IP that is sold before development.

Closing Thoughts

“In sum, these questions ask whether it is contradictory to support Canadian IP development, retention, and commercialization while keeping Canada an attractive place to do business for international players. This paper contends that both can be supported via a strengthened Canadian innovation ecosystem, with attention to business density, entrepreneurial experience, improved capitalization, and better IP commercialization literacy for Canadian researchers and entrepreneurs.”

FDI has its place in development of any economy. On the other hand, the stage and means by which FDI is incorporated in the IP development pipeline is a key consideration that is the difference between a thriving deep tech sector and being an IP republic. Given Canada’s lagging innovation metrics and productivity, the current frameworks governing IP arising from publicly-funded research are off the mark.

Canada has all the ingredients it needs to excel in technology commercialization, but as the report clearly states, there is a need to address underlying structural barriers and recognize that individual elements of the innovation pipeline form a complex interacting system and cannot be considered in isolation. The Matthews and Rice report is an excellent synthesis of the challenges and opportunities, as well as approaches that can and should be considered in addressing Canada’s innovation woes.

My compliments to the ICTC team for their excellent work.