Canada Doesn’t Have an Innovation Problem. It Has a Scale Problem.

The industries of the future are booming — just not in Canada.

South of the border, the United States continues to produce new technology giants and frontier startups. American artificial intelligence firms are scaling rapidly, supported by deep venture capital markets and dense talent networks in places like Silicon Valley and Boston.

The story is different in Canada. Canada produces fewer startups per capita, pays lower technology wages, attracts less business investment and increasingly watches its most promising firms relocate elsewhere.

Advanced industries are more productive. They generate higher wages and generate positive spillovers for the rest of the economy. When a country fails to build and retain them at scale, the effects ripple through the broader economy.

Canada’s productivity shortfall is well documented, and politicians, economists and business leaders all offer competing explanations. Most solutions converge on familiar tropes: liberalize the economy, increase competition, and reduce government spending to lower overall tax burdens. These ideas, while helpful, are insufficient. 

Canada already ranks ahead of the United States on many measures of economic freedom. Before the Tax Cuts and Jobs Act (2017), Canada’s effective corporate tax burden was also significantly lower. If taxes and regulation were the decisive explanation, Canada’s productivity gap would have narrowed rather than widened over the past two decades.

Moreover, the states that lead in frontier industry, such as California and Massachusetts, face high regulatory, tax, and cost-of-living burdens that often exceed those of Canada. Yet, they remain the most innovative economies in the world.

The Missing Ingredient: Industrial Clusters

The missing piece is not government-imposed barriers, but the scale and concentration of industry — what economists refer to as agglomeration. Innovation is not evenly distributed. It clusters together tightly. Places like Silicon Valley, Boston and Shenzhen form dense, self-sustaining, industry-specific ecosystems that become far more productive than the sum of their individual firms.

Agglomeration effects arise when talented workers and ambitious firms pack together. Workers move between companies, spreading new ideas about how to make things better and faster. Firms compete aggressively for talent, bringing the smartest and most capable people together in one place. Capital is plentiful and circulates quickly, as both investors and liquidity-seeking firms solicit one another. The end result is that there is intense competition and scale in one place, driving productivity.

Even when policy environments are imperfect, such as the case of high-tax, high-regulation California, the ecosystem is capable of sustaining momentum due to its preexisting characteristics. Technology firms have a strong incentive to stay at the cutting edge of R&D, and that means they need access to the best and brightest workers. Access to talent outweighs the tax advantages offered by jurisdictions with weaker innovation ecosystems, so firms stay put despite higher costs.

Likewise, the workers themselves want the highest wages from the leading companies, and a software engineer looking outside of San Francisco will find fewer opportunities at a lower after-tax income. No one wants to take the cut and leave, so no one does. Both the worker and the firm will tolerate higher costs, including paying large tax sums to the state government,  because losing access to each other would be even more costly.

A strong cluster doesn’t just attract talent and capital; it also insulates them from poor policy choices. Strong industrial clusters can continue to outperform even in policy environments that many economists would consider less competitive.

One influential paper found that a one percentage point increase in the personal tax rate reduces patents by 0.7–0.8% in U.S. jurisdictions where there was a dense technological cluster in that field, versus 1.1–1.2% elsewhere. Countries with more agglomeration and economies of scale are also more likely to retain human capital. In one study of high-performing “superstar inventors”, higher taxes on the wealthy strongly incentivized them to relocate internationally, but this fell substantially when the firms they worked for already had a strong presence anchoring them in a country. That means for countries like Canada or the United Kingdom, the elasticity response of superstar inventors to higher taxes was 0.31 and 0.36, respectively; for the United States, it was only 0.003. Homegrown talent is roughly a hundred times more likely to be taxed out of Ottawa or London than out of Silicon Valley.

Why Canada Struggles to Scale

By contrast, Canada does not have any comparable superstar cities.

The Toronto-Waterloo Corridor is our closest analogue to Silicon Valley. It does have world-class talent and impressive firms. But it has also struggled to generate the scale and self-reinforcing dynamics seen in the U.S.

Canada is caught in the same dynamic that makes global tech hubs so big, just on the losing end of it. A cluster below critical mass can’t hold its own people, and the incentive structure that keeps a Californian engineer in San Francisco runs the other way for a Torontonian, drawing them abroad.

Canada produces strong STEM graduates, but a significant share leave for the United States, lured by the promise of higher tech wages. Canada also struggles to retain its most promising firms and innovators, which want to be where they can easily access talent and investors. A recent survey of nearly 3,000 Canadian‑founded “high‑potential” startups found that only about 32 % were headquartered in Canada in 2024, down from over 67 % a decade earlier, with the majority of the remainder based in the United States.

Why Scale Matters

The first problem is scale. Statistics Canada estimates that roughly 60% of the US-Canada productivity gap since 2000 reflects Canadian firms being too small. Small firms can’t sustain the capital depth or specialized labour pools that the frontier demands, causing human capital to leave for larger, more efficient firms that offer higher returns (and aren’t in Canada).

Another portion of the gap is a lack of business investment in physical capital. In Canada, there is less money available to invest in growing firms, and our firms generally offer lower returns for each dollar invested. Canadian founders in the U.S. raise almost twice as much as those in Canada, per the same survey. Without a policy fix, we will continue to see Canadian success stories, like Element AI, exit the country the moment they’re ready to go big.

Compounding this, nearly 40 percent of investment flows into real estate — the highest share in the OECD. Capital that might have funded globally competitive firms is instead locked into ever-more expensive real estate. Failing to address housing affordability also risks constraining growth in Canada’s most productive regions, preventing the proximity and density that are essential to innovation.

Countries that host frontier industries are often the ones that shape global standards, attracting strategic investment and exercising greater geopolitical influence. In an era where artificial intelligence, semiconductors and advanced manufacturing are becoming strategic assets – and, increasingly, geopolitical leverage – Canada’s inability is no longer an exclusively economic problem. It is a question of national sovereignty.

Building the Conditions for Growth

Canada’s challenge is one of scale rather than an inherent lack of entrepreneurship. Policy should focus more on helping promising firms become globally competitive businesses, over generating many small ones.

Canada’s startup support system, where it currently exists, is biased towards smaller firms at earlier stages of development. Size-based tax preferences, including the Small Business Deduction, phase out as firms grow. Meanwhile, the Scientific Research and Experimental Development (SR&ED) tax credits disproportionately reward laboratory research and technical uncertainty rather than commercialization or scaling. By contrast, the United States provides full expensing for capital investments by firms, significantly incentivizing business investment there. 

Low taxes and light regulation have not, on their own, conjured a frontier cluster anywhere. The most market-friendly American states have not become innovation hubs, and the few conservative states that are hubs, like Texas, didn’t get there solely by deregulating. 

Texas offers a useful counterexample. It did not become a technology hub through deregulation alone. Public universities produced scientific talent, NASA contracts likely helped to anchor the aerospace sector in Houston, and the federally-funded SEMATECH consortium helped establish Austin as a national semiconductor centre.

More ambitious interventions may then be needed. China, for example, sports industrial clusters that are several cities wide, all supported via aggressive government fiat. 

Policymakers should be cautious here. The evidence for large-scale industrial policy is context-dependent and somewhat mixed, with horror stories of expensive boondogles contrasted with provable success stories. What works and what doesn’t is not always immediately clear.

But no one in Canada is proposing a Five Year Plan, nor should they. Canada needs to prove it can focus support where it’s needed before a moonshot. Canada substantially increased its population over the last five years to achieve scale, but much of this increase was focused on low-skilled temporary workers, French-speakers and students in lower-ranked and two-year degree programs, rather than focusing on recruiting top talent where it’s needed. Efforts at industrial development frequently take the form of piecemeal subsidies to specific firms – like Bombardier, Volkswagen, and Northvolt, rather than broad-based sector support or incentives.

And if the time comes to explore more aggressive options, economic evidence suggests it should favour upstream industries, as government support to industries at the beginning of the supply chain is more likely to filter down to industries as at the very end of it.

The industries of the future will not wait. If Canada wants them built here, it needs the conditions necessary for them to flourish. That may require more than incremental liberalization, and perhaps more than the immediate solutions offered here. In the end, however, the cost of inaction may be higher. Canada’s productivity continues to decline relative to the U.S. Without correction, Canada will continue ceding the industries of tomorrow, and the economic power they bring, to other countries.