Reviewing Budget 2024 Science and Innovation Funding
Boosting research output and industrial patent investment without a bridge between them is unlikely to result in long-term economic growth
While most of the Canadian innovation ecosystem focuses on the capital gains tax, I took a close look at the rest of the innovation-focused budget initiatives in Budget 2024. Specifically, this article examines Section 4.1, focusing on proposed budget items that relate to research, intellectual property generation, and early-stage economic development of the results of both.
The intro to Chapter 4 starts strong, with a statement Iām sure we can all get behind.
āTo ensure every Canadian succeeds in the 21st century, we must grow our economy to be more innovative and productive. One where every Canadian can reach their full potential, where every entrepreneur has the tools they need to grow their business, and where hard work pays off.ā
From there, it quickly becomes clear that the core lesson has not yet been learned: boosting scientific productivity does not lead to downstream economic activity without a bridge to retain and continue development of research outputs. The budget has significant investments in science and anticipated changes to how grants are administered, but fails to lay out a viable plan to retain the results, or to bridge the anticipated outputs from lab to market.
Foreign Direct Investment
ā[ā¦] the government's economic plan is investing in the technologies, incentives, and supports critical to increasing productivity, fostering innovation, and attracting more private investment to Canada. This is how we'll build an economy that unlocks new pathways for every generation to earn their fair share.ā
While the paragraph above does not specifically mention foreign private investment, charts 4.1 and 4.2 make clear that this is what is meant. For reason that remain genuinely unfathomable to me after 150 years of evidence to the contrary, the government still equates foreign direct investment (FDI) in Canada with innovation mission success. (Note carefully that I refer to the Canadian government as a general entity, not specifically this government, given the longevity of the problem. This is an ongoing and bipartisan failure that has survived many changes of political control).
Chart 4.2, reproduced above, proudly announces that Canada is the beneficiary of the third-most foreign direct investment (FDI), completely misunderstanding that this is part of the problem. FDI is only of value if it is done in a way that builds up domestic production capacity in the long term. FDI has a place in economic development, but it is not at the IP generation stage. Because of this misunderstanding, Canada is and has for decades been effectively an IP republic, bringing greater benefit to the rest of the world than to the taxpayers footing the bill. Most of the initiatives laid out in Section 4.1 of Budget 2024 fall short of fixing it.
Artificial Intelligence
āCanada's artificial intelligence (AI) ecosystem is among the best in the world. SinceĀ 2017, the government has invested over $2 billion towards AI in Canada. Fuelled by those investments, Canada is globally recognized for strong AI talent, research, and its AIĀ sector.
Today, Canada's AI sector is ranked first in the world for growth of women in AI, and first in the G7 for year-over-year growth of AI talent. Every year sinceĀ 2019, Canada has published the most AI-related papers, per capita, in the G7. Our AI firms are filing patents at three times the average rate in the G7 [ā¦]ā
There are several fallacies here. Letās unpack them.
While Canadaās AI ecosystem was among the best in the world in 2017, we squandered that lead by waiting too long to build on it. While we are in the top 5 countries for investment in AI in the last 5 years, the gap between first and second, and between second a third, makes any ranking beyond that irrelevant. Our investment is about par with the rest of the western world and is unlikely to move the needle globally.
By itself that is not overly surprising nor is it an issue: Canada is a relatively small, open economy and there was never any expectation that our investment will match the US or China in anything technology-related, let alone AI. The issue is that while we had a thought leadership position in AI policy development through The Montreal Declaration for a Responsible Development of Artificial Intelligence in 2017, we did nothing to build on that at a national level, delaying any commitment on regulation past the point where it would have had any value as an input to the global conversation. Canada has never had a hope of leading the AI race in output, but we did have an opportunity to lead the world in guiding how that race was conducted. That boat has sailed.
The statement that ā Canada has published the most AI-related papers, per capita, in the G7. Our AI firms are filing patents at three times the average rate in the G7ā is itself problematic because it leaves out any mention of what happens IP after it is generated. According to a small-scale study by Canadian IP leader Jim Hinton, as much as 75% of Canadaās AI-related IP leaves the country via the big tech firms and 18% collects dust in University tech transfer offices, while only 7% is actually being used in the Canadian private sector. The licensing fees collected from patents that leave the country are miniscule compared to the cost of development.
While investment in domestic compute capacity is a good thing, we need to take control of the output of that process before we can hope to benefit from having the infrastructure to accelerate our R&D productivity. If more than 90% of our AI-related patent output is accruing no benefit to Canada, and 75% of it is contributing to advancing other countries, then the minority of AI-related IP that actually makes it into the Canadian private cannot possibly yield enough economic growth to justify any level of investment. $2B for AI would be an excellent idea if we had the capacity to retain, incubate, and develop the IP that will result. Without first plugging the IP leak, Canada will not reap the benefits.
Incentivizing More Innovation and Productivity
āTo incentivize investment in innovation-enabling and productivity-enhancing assets, BudgetĀ 2024 proposes to allow businesses to immediately write off the full cost of investments in patents, data network infrastructure equipment, computers, and other data processing equipment. Eligible investments, as specified in the relevant capital cost allowance classes, must be acquired and put in use on or after Budget Day and before JanuaryĀ 1, 2027. The cost of this measure is estimated at $725Ā million over five years, starting inĀ 2024-25.ā
This is a sound idea, taken in isolation. Patent costs are onerous, particularly in the early days of development, and anything that can reduce the cost burden of intellectual property retention on companies that are actually developing them into products and services is a positive step.
On the other hand, when combined with the massive disincentive to invest in those same startups represented by the capitals gains inclusion rate increase (this is the only time I will bring it up in this article, I promise), any impact of this measure will be washed out. Companies that are developing the patents coming out of Canadian research institutions need capital to get to the point of having patent costs to write off in the first place, and the capital gains change will do nothing to help with the process of finding it.
If combined with an incentive for private sector investment in the same companies that are able to write off patent costs alongside updates to SR&ED and a patent box implementation, this measure could be a powerful motivator for domestic IP retention. As things stand, I expect the potential impact of this write-off to be diluted by the lost access to development capital.
The fact that the tax incentive discussed here focuses entirely on industry patent costs reinforces the idea that the government does not understand the reasons behind a lack of IP retention and commercialization. First, a large fraction of the IP that Canada produces originate in our academic institutions, not our private sector, and it is via academia that much of it leaves the country. Second, even in the minority of cases where Canadian private industry invests in a patent by licensing it from academia, investment in a patent does not necessarily mean economic development. Many patents licensed to industry are licensed with the intention of burying them in an IP moat, rather than developing them. Existence of retained patents is not enough. The government needs to develop metrics relating to productive use of IP and use them as prerequisites for related tax breaks.
Boosting R&D and Intellectual Property Retention
āResearch and development (R&D) is a key driver of productivity and growth. Made-in-Canada innovations meaningfully increase our gross domestic product (GDP) per capita, create good-paying jobs, and secure Canada's position as a world-leading advanced economy.ā
Given the point of this article so far, one would think that a heading relating to IP retention would be cause for celebration, but unfortunately it simply reinforces the point that the government misunderstands the dynamics that cause IP to leave.
The key point that is being consistently missed is that by the time these measures kick in, a significant fraction of IP is already gone. SR&ED focuses on industrial R&D but stops short of even patenting (note that I did not say domestic R&D; historically most of the credit goes to multinationals with Canadian branch plants), while the patent box only kicks in once revenues occur, which could be delayed from the point of IP generation by years.
Both measures require that IP make it out of academia, into a domestic firm, and then actually begin the development phase in the private sector before either incentive provides any retention pressure. With so much of the IP developed in academia being licensed away before this phase, the impact of these measures will be diluted compared to their potential without first building the bridge between academic and industrial IP development. The current bridge, in the form of academic tech transfer offices, are insufficiently resourced and empowered to change this, and the results speak for themselves.
Research Support
āCanada's granting councils already do excellent work within their areas of expertise, but more needs to be done to maximize their effect. The improvements we are making today, following extensive consultations including with the Advisory Panel on the Federal Research Support System, will strengthen and modernize Canada's federal research support.ā
The issues that I am raising are not new. The quote above references the Bouchard Report, which focused on the Canadian research landscape, identifying several key issues with how it is conducted currently. In particular, the report accurately found that the current approach of the tri-council agencies, among other things,
limits opportunities to foster pre-commercialization activities and link them to commercialization activities; and
inhibits the capacity to set forward-looking strategic direction for science, research and innovation
A detailed review of the Bouchard Report will be the subject of a future post, but I mention it here since Budget 2024 proposes to follow up on one of their recommendations: create a capstone organization that can better coordinate the Canadian research funding and activity.
While these budget measures could have a positive impact, they remind of the Canadian Innovation Corporation (CIC), a much-needed overhaul of NRC IRAP that has been repeatedly delayed to the point that it now has to survive what promises to be a disruptive election cycle if it is to ever exist at all. Given the lack of detail on how this capstone organization would work, and the distinct possibility that it goes the way of the CIC, I will reserve judgement for the time being.
I will, however, provide some concrete advice to those tasked with designing this organization, should it ever come to fruition: the metrics become the targets. If the impact of science funding is measured in papers published and patents filed, then that is where the process will stop, and research that ends in the lab is of no economic use to Canada.
By far the most important design consideration for the capstone organization is how success will be measured, and it can take a long time for impact to be felt. It will be critical to design the KPIs to reflect actual target outcomes (long-term economic development of technology and flow of IP resulting from Canadian research funding) rather than intermediate proxies like publications and patents.
Done well, this capstone organization could be the bridge between academia and industry that we so badly need. Done poorly, and it only add another layer of bureaucracy on top of an already bloated system.
Talent Retention
āTo build a world-leading, innovative economy, and improve our productive capacity, the hard work of top talent must pay off; we must incentivize our top talent to stay here.
Federal support for master's, doctoral, and post-doctoral students and fellows has created new research opportunities for the next generation of scientific talent. Opportunities to conduct world-leading research are critical for growing our economy. In the knowledge economy, the global market for these ideas is highly competitive and we need to make sure talented people have the right incentives to do their groundbreaking research here in Canada.ā
I am happy to see the proposal to increase student scholarship amounts, though disappointed at both the amount and to see a lack of commitment to indexing these to inflation on an ongoing basis. The increase of masterās scholarship from $27,000 to $40,000 only reflects catching up to inflation since 2010, meaning that these scholarships have been declining in real value steadily for more than a decade. Even after this change, Canadian graduate students with scholarships will be significantly worse off now than they were 10 years ago in Canada.
These measures are better than nothing, and long overdue, but hardly qualify as a strong push to retain talent. Paying grad students more and giving out more scholarships may result in more grad students being trained in Canada, but it does nothing for retention past that point. Just as with technologies and patent portfolios, Canada needs to give some careful thought to bridging the talent that this funding will train into the domestic workforce. It is an unfortunate but increasingly obvious reality that the vast majority of grad students will never be academic faculty. The rest will turn to industry, and unless Canada has related industrial development, they will leave, and they will be right to do so.
Looking Forward
āCanada's skilled hands and brilliant minds are our greatest resource. Capitalizing on their ideas, innovations, and hard work is an essential way to keep our place at the forefront of the world's advanced economies.ā
The sooner we start putting those words into action, the better, but the plans laid out in Section 4.1 demonstrate that key lessons have not yet been learned. The individual initiatives suggested by Budget 2024 are mostly well-intended, and would likely be quite effective if issues of IP and talent retention had already been addressed. However, this part of the budget fails to recognize that none of the results (IP or talent) will stay in Canada and that much of the proposed spending will not benefit the Canadian economy in the long term.
Our focus should be on fixing the leaks and building bridges between research and early economic development before investing in either activity in isolation. As presented, the plan appears to be throwing money at the wall and hoping (without even checking) that some of its sticks.