Capital Stock per Worker

Canada vs US capital per worker: the 55-cent problem

If you want one chart that explains why Canadian wages have stalled and why the Canada–US labour productivity gap keeps widening, this is it. The C.D. Howe Institute's December 2025 Commentary No. 699 — "Canada's Investment Crisis" — found that for every dollar of new capital a US worker received in 2025, a Canadian worker received roughly 55 cents, with similar deficits against the broader OECD average. Director of Research William Robson and his co-authors call it a "vicious cycle": underinvestment lowers productivity, which lowers profits and wages, which discourages further investment.

For the indicator definition (capital stock per employed person, OECD 2020 PPP basis) see the Capital per Worker indicator page. This page is about the Canada–US gap — and it is a story of a crossover most Canadians do not know happened.

The numbers

From the OECD index series shown above (2015 = 100):

• Canada, 2000: about 67.05
• United States, 2000: about 62.73 — Canada was ahead at the turn of the century
• Canada, 2022: about 102.73
• United States, 2022: about 114.22 — and now the US is ahead by roughly 11 points

The crossover happened around 2010 and the gap has widened more or less continuously ever since. In ratio terms, Canadian capital per worker is now growing at less than half the US pace, and the C.D. Howe analysis breaks the deficit out further by asset type:

• Machinery and equipment: Canadian workers receive about 41 cents of new investment for every US dollar.
• Intellectual property products (software, R&D, designs): about 32 cents on the US dollar.
• Non-residential structures: closer to parity but still trailing.

The composition matters. The categories where Canada is furthest behind — software, R&D, and machine tools — are exactly the categories that drive measured productivity growth in modern advanced economies. Canada is investing in the wrong things, and not enough of them.

How a leader became a laggard

Canada's pre-2010 lead was largely a story about the resource sector. The 2003–2014 commodity supercycle drove enormous capital expenditure in Alberta's oil sands, in pipelines, and in mining infrastructure. Statistics Canada's December 2025 productivity review notes that capital deepening in extractive industries was the single largest contributor to the Canadian capital stock through that period.

When WTI crude collapsed from roughly $107/bbl in June 2014 to about $44.50/bbl in January 2015, Canadian oil and gas capital expenditure fell from a record CAD $76.1 billion in 2014 to about $37.8 billion by 2016 — more than halved in two years. That alone removed about two percentage points from Canadian GDP growth in 2015. Trevor Tombe at the University of Calgary has called this period the start of "the Great Divergence" — see GDP per capita for the per-capita version of the same story.

The expectation in 2016 was that other sectors would pick up the slack. They did not. Non-energy business investment in Canada has been flat or declining ever since, while US business investment — particularly in software, semiconductors, and AI infrastructure — has accelerated.

The US side: an AI capex supercycle

The US lead is widening because of where Americans are putting their capital. Calendar 2025 saw US hyperscaler capital expenditure reach approximately $443 billion, with consensus forecasts pointing to roughly $600 billion for 2026. The bulk of that flows to data centres, GPU clusters, advanced cooling, transmission upgrades, and the fabrication facilities that supply them.

The Information Technology and Innovation Foundation (ITIF) noted in 2025 that US business investment in IT equipment per worker is now several multiples of the Canadian equivalent, and that the gap is expanding rather than contracting. Microsoft's announcement of a $19 billion Canadian cloud and AI infrastructure investment, made in 2025, is meaningful but small relative to what is happening south of the border. Canada is on the receiving end of foreign capex; the US is the foreign capex.

Why this matters for wages

The mechanism is mechanical: real wages over long periods track the marginal product of labour, and the marginal product of labour is largely a function of the capital and technology each worker has to work with. If the average US worker has access to roughly twice the new capital each year, the long-run real wage trajectory is going to diverge in a way no labour-market policy can offset.

Senior Deputy Governor Carolyn Rogers's now-famous March 26, 2024 "break the glass" speech in Halifax framed the productivity problem in exactly these terms — see the labour productivity comparison for the full quote and the policy fallout. Q4 2025 real GDP came in at −0.2% quarter-over-quarter, and Statistics Canada's small-firm productivity paper found that Canadian SMEs are roughly 70% as productive as US SMEs largely because they have less capital to deploy.

Carney and Champagne's 2025 response

Unlike earlier in the decade, Canadian fiscal policy has actually moved on the investment file:

• Productivity Super-Deduction, announced by Finance Minister François-Philippe Champagne in Budget 2025 (November 4, 2025): a package of accelerated depreciation measures designed to bring the marginal effective tax rate on new business investment from 15.6% to 13.2%, which would make it the lowest in the G7.
• Bill C-5, the One Canadian Economy Act (Royal Assent June 26, 2025): legislation aimed at breaking down interprovincial trade barriers, with IMF estimates suggesting full internal trade liberalization could raise long-run real GDP by approximately 7%.
• Prime Minister Mark Carney's stated CAD $1 trillion target for total investment over five years, including roughly CAD $280 billion in direct federal capital and incentives intended to mobilize the rest from private and institutional investors.

Whether these measures actually move the chart is an empirical question. C.D. Howe's verdict in late 2025 was cautious: the super-deduction is meaningful but not large enough on its own to close a gap that has been widening for fifteen years. Robson's published view is that without sustained, multi-year acceleration in business investment, the chart simply does not bend.

What to watch in 2026

1. Quarterly non-residential business investment in the Statistics Canada GDP release. This is the highest-frequency leading indicator.
2. Software and IP investment specifically — the categories where Canada is most behind and where the largest US gains are concentrated.
3. C.D. Howe's annual capital-per-worker update, which is the most cited single benchmark in Canadian policy debates.
4. Carney's $1 trillion target progress reports — credibility will depend on whether early-year capital flows match the announced trajectory.
5. US AI capex moderation. If the hyperscaler capex cycle slows in late 2026, the gap could stabilize for mechanical reasons rather than Canadian improvement.

The stakes

Capital per worker is the single most important determinant of long-run real wages, and Canadian capital per worker is growing slowly enough that the GDP per capita gap and the productivity gap are both downstream of this chart. The crossover happened around 2010, the divergence accelerated after the 2014 oil shock, and the 2025 policy response is the most ambitious in a generation but still smaller than the underlying trend it has to bend.

The short version: Canada was ahead in 2000, fell behind by 2010, and is now investing roughly 55 cents per worker for every US dollar. Closing that gap is the central challenge of Canadian economic policy in the late 2020s, and the chart above is the scoreboard.