Win the Global Talent War

Summary

Canada spends more on university research than almost any peer country, but gets far less out of it. Discoveries die in the lab. Patents flow abroad. One in four STEM graduates from top Canadian universities works outside the country. Leading entrepreneurs leave and are drawn to places with capital formation density that can support their career paths.  

The problem isn't a lack of talent or ideas — it's that for decades, Canada has had no functioning bridge between research and revenue. While early stage capital formation has improved, we do not have sufficient capital formation to support scaling our champions.

Meanwhile, 53% of all venture capital invested in Canada comes from the United States, making our innovation economy dangerously dependent on foreign decision-makers. This memo proposes three fixes: rewire university technology transfer to reward commercialization, create domestic capital incentives so startups can scale here, and launch a sovereign talent fund to recruit and retain the people who will build Canada's future.

‍

Narrative

A Canadian PhD student earns about C$24,000 a year.¹ A PhD student doing the same work at an American university earns between C$52,000 and $65,000.² Canada tells its best young scientists: we'll train you, but we won't pay you. Graduate stipends in Canada stayed frozen for over 20 years before Budget 2024 finally raised them.³ While places like Perimeter institute that have both public and private money pay their PhD students in Physics more on average, they still fall behind competing schools in the US. It is no surprise that the majority of current graduate students surveyed by the Ottawa Science Policy Network are considering leaving Canada after finishing their degrees.⁴

Canada does many things right. It is the home of some of the best universities in the world. The University of Toronto, UBC, and McGill all rank in the global top 50. Canada has about 4% of the world's top 500 universities despite being just 0.5% of the global population.⁵ Federal research funding got a meaningful boost in Budget 2024, with $1.8 billion over five years for core research grants.⁶ 

But the pipeline from research to economic value is broken. Canada has one of the highest rates of higher education R&D spending in the developed world — nearly double the rate of the U.S. and South Korea as a share of GDP.⁷ Yet it gets far less commercial output per dollar. According to AUTM data, American university tech transfer offices license patents at three times the rate of Canadian ones.⁸ Canada's overall R&D intensity sits at 1.81% of GDP, well below the OECD average of 2.73%,⁹ and business R&D spending has been falling for over two decades even as the OECD average has risen. Deep learning, the technology behind ChatGPT, self-driving cars, and virtually every AI product in the world, was invented in Canadian labs, funded by Canadian taxpayers, at the University of Toronto and Université de Montréal. Today, the companies built on that technology — Nvidia, Google, Meta, Microsoft, OpenAI — are collectively worth over $15 trillion. AI is projected to add $15.7 trillion to the global economy by 2030.¹⁰ Canada's share of that value is a rounding error. The research was ours, but the wealth is theirs.

The commercialization gap means discoveries made with Canadian taxpayer money end up owned by foreign firms. — and the profits go with them. The United States alone earned US$128 billion in 2022 from selling intellectual property abroad¹¹ — much of it built on research and talent that originated in countries like Canada. An IRPP study found that 45% of the 9,000-plus patents granted by the U.S. Patent Office in 2017 with Canadian inventors were immediately assigned to entities outside Canada¹² — a share that more than doubled over 20 years. Of those that stayed in Canadian hands, many were later sold abroad. Foreigners control one-third of all in-house R&D spending in Canada — $11.5 billion of the $30.4 billion total. Canada is subsidizing other countries' innovation.¹³

In 2024, 53% of all venture capital invested in Canadian companies came from U.S. sources.¹⁴ In Q1 2025, Canadian private VC firms invested just $95 million — while U.S. investors put in $612 million.¹⁵ When the U.S. provides most of your startup capital, American investors set the terms. Those terms often mean: move your headquarters to Delaware, relocate your engineering team, and sell when they say it is okay to sell, often to a U.S. acquirer. In 2024, there were zero VC-backed IPOs in Canada.¹⁶ The exit market is effectively shut.

Meanwhile, the talent drain accelerates. The Conference Board of Canada and Institute for Canadian Citizenship's "Leaky Bucket 2025" report found that one in five immigrants leaves Canada within 25 years, with highly educated immigrants departing at twice the rate of those with less education. Doctorate-holders are nearly twice as likely to leave as those with a bachelor's degree. About 828,000 Canadian-born immigrants lived in the United States as of 2023,¹⁶ and a 2018 study found one in four STEM graduates from U of T, UBC, and Waterloo were working outside Canada, with 66% of software engineering graduates leaving.¹⁷

Other countries don't wait for talent to show up, they invest, attract and give them opportunities to prosper and contribute to the economy. Canada should do the same.

‍

Existing Solutions

United States — The Bayh-Dole Act. Before 1980, the U.S. government held 28,000 patents from federally funded research and fewer than 5% were commercially licensed.¹⁸ The Bayh-Dole Act let universities own and commercialize inventions from public research. Since then, it has contributed to over $1.3 trillion in economic growth, 4.2 million jobs, and 11,000 new startup companies.¹⁹ Canada has no equivalent legislation.

Denmark — Industrial PhD Program. Denmark's program, run by Innovation Fund Denmark since 1971, embeds PhD students inside companies while they complete their degrees. The government co-funds the salary. Results: nearly 50% of participating companies have obtained a new patent,²⁰ and the program is linked to higher revenue, employment, and productivity for firms. It directly solves the gap between academic research and commercial application.

United Kingdom — Global Talent Fund. UKRI's Global Talent Fund awards institutional grants to 12 top UK research organizations to recruit international researchers.²¹ Funding covers research costs, visa fees, and relocation for researchers and their families across industrial strategy priority areas. It's a direct, aggressive investment in poaching the world's best talent.

‍

What Must Be Done

Canada doesn't need more reports. It needs five things: turn university research into Canadian companies, keep taxpayer-funded inventions in Canadian hands, build domestic capital so startups can scale here, pay the world's best scientists to move here, and fix immigration to select for the skills we actually need.

1.Tie 20% of tri-council funding to commercialization outcomes, starting with NSERC — and extend the same requirement to Canada's national AI institutes. Budget 2024 gave the tri-council agencies $1.8 billion over five years in new core funding. None of it is tied to whether research becomes a Canadian company. Change that. Amend the NSERC funding framework so that 20% of grant renewals are scored on commercialization metrics: Canadian startups formed from research, domestic patent licenses executed, and industry co-funding secured. Model the scoring on Denmark's Industrial PhD Program, where Innovation Fund Denmark tracks patents, revenue, and employment from every project. Apply this to NSERC first. Expand to CIHR by 2028. Universities that produce founders and licensees get more funding. Those that don't, get less. Apply the same logic to Canada's national AI institutes. Since 2017, the government has poured over $2.4 billion into AI through the Pan-Canadian AI Strategy and Budget 2024, plus another $925 million in Budget 2025. Yet 75% of patents produced by Canada's two leading AI institutes leave the country, and only 6.1% of Canadian businesses were using AI as of mid-2024. Canada invented deep learning. Every dollar of public AI funding should come with a commercialization scorecard: Canadian startups created, IP retained domestically, and Canadian companies adopting the technology. Institutes that can't show results within three years should see their funding reallocated to those that can.

2. Pass a Canadian Innovation Ownership Act. The U.S. Bayh-Dole Act gave universities the right to own patents from federally funded research. Before it passed, fewer than 5% of 28,000 government-held patents were licensed. Canada has the same problem — 45% of Canadian-invented patents are assigned to foreign entities at issue. A Canadian version should give universities title to tri-council-funded IP, require a 24-month domestic-first licensing window, and mandate annual commercialization reporting to the Treasury Board.

3. Create a Canadian Growth Capital Tax Credit. In 2024, 53% of VC dollars in Canada came from the U.S. In Q1 2025, Canadian private VC firms invested just $95 million — less than one-sixth of U.S. contributions. Offer a 25% refundable tax credit to Canadian pension funds, insurers, and family offices on the first $500 million invested annually in Canadian-headquartered companies at Series A and beyond. Require investees to maintain Canadian HQ and majority-Canadian boards for five years. Fund it by eliminating the LSVCC tax credit, which costs over $100 million per year and consistently underperforms. Complement this with a Canadian Sovereign Wealth Fund — seeded with $3 billion in federal capital and co-invested alongside the private sector — to anchor late-stage rounds in strategic sectors such as AI, clean tech, and advanced manufacturing. This dual structure — private incentive plus public anchor — will crowd in domestic and foreign capital while retaining strategic industries at home.

4. Launch a Sovereign Talent Fund. Pay $100,000 signing bonuses to 1,000 top researchers and engineers per year. BCG has modelled this: a fund attracting 1,000 skilled workers requires $15 million to launch, becomes self-funding within 20 months, and generates $180 million in net fiscal value within four years. Budget: $100 million per year, funded by reallocating 2.5% of IRCC's $4 billion settlement budget. Run it through a new independent agency with no ministerial approval for individual selections. Selectees get permanent residency within 90 days.

5. Double down on STEM immigration — and recognize foreign credentials. IRCC introduced STEM selection in Express Entry in 2023, but held only one STEM-specific draw in 2024. Commit to at least six STEM draws per year and lower the CRS threshold to 420. Add a 50-point CRS bonus for PhDs from top-100 global universities in STEM. But getting people here is only half the problem. Over 25% of immigrants with foreign degrees work jobs requiring only a high school diploma. Engineers drive taxis. Doctors work as assistants for years. Automatically recognize credentials from the top 10 GDP countries for shortage occupations, with a 90-day bridging assessment instead of years-long relicensing.

‍

Common Questions

"Aren't you just poaching talent from developing countries?" Research shows two-way knowledge flows between origin and destination countries. The people most likely to move are already globally mobile. They're choosing between Canada, the U.S., and Europe. Canada isn't taking talent from vulnerable nations, it's losing a competition for people who have options. Doing nothing means they go to the U.S. instead.

"Won't universities resist tying funding to commercialization?" Some will. But Canadian universities have had 20 years to improve tech transfer and the results are poor — U.S. universities license patents at three times the rate of Canadian ones.²³ Universities receive billions in public funding, taxpayers deserve a return. 

"$100 million a year for 1,000 people? That's expensive." It's one of the best investments a government can make. BCG's modelling shows a talent fund of this size generates $180 million in net fiscal value within four years.²⁴ A single top AI researcher can anchor a lab that creates dozens of jobs and millions in economic output. Canada already spends over $4 billion a year on immigration settlement. 

‍

Conclusion

Canada has world-class universities, smart people, and good research. What it doesn't have is a system that turns those strengths into Canadian companies, Canadian jobs, and Canadian wealth. The pipeline is broken. The capital is foreign. The talent is leaving. Every year Canada waits, more patents go abroad, more startups get acquired by American firms, and more of our best people take their skills south. Fix the commercialization pipeline. Build domestic capital. Go get the world's best talent and make it worth their while to build here. This isn't a nice-to-have. It's the difference between a country that invents the future and one that watches it happen from next door.