Central bankers are trained to speak in hedged, carefully modulated language, because markets move on their adjectives. So when the second-in-command at the Bank of Canada stood in front of a business audience in Halifax and reached for the metaphor of breaking emergency glass, it was not rhetorical flourish. It was a deliberate decision to spend institutional credibility on getting attention. Two and a half years later, it is worth asking what the business community actually did with it.
Key Takeaway
On March 26, 2024, Bank of Canada Senior Deputy Governor Carolyn Rogers told the Halifax Partnership that Canada's productivity record had become an emergency, noting that the level of productivity in Canada's business sector was more or less unchanged from where it had been seven years earlier, following six consecutive quarters of decline. She attributed this to weak business investment in machinery, equipment and intellectual property, labour market matching failures, limited competition, regulatory uncertainty, and a risk-averse business culture, and noted that Canada has a larger proportion of small businesses, which tend to invest less. By September 2025 the C.D. Howe Institute reported the situation had worsened: real machinery and equipment investment fell nine percent in the second quarter, leaving it twelve percent below its level a year earlier, with output and incomes per Canadian set to fall in 2025 for the third consecutive year. The most actionable finding in the entire file is buried in the coverage: few employers actually measure productivity at all.
The Speech
The remarks were delivered by Carolyn Rogers, Senior Deputy Governor of the Bank of Canada, to the Halifax Partnership in Nova Scotia on 26 March 2024, under the title "Time to break the glass: Fixing Canada's productivity problem"[1].
The framing was explicit. As Rogers put it: "I want to talk about Canada's long-standing, poor record on productivity and show you just how big the problem is. You've seen those signs that say, 'In emergency, break glass'. Well, it's time to break the glass"[2]. The Globe and Mail described the speech as unusually blunt, reporting Rogers saying "I'm saying that it's an emergency, it's time to break the glass" to the business audience[3].
Rogers chose her venue deliberately: an organization dedicated to helping Canadian businesses grow and thrive[4]. The message was aimed at business, not at government alone, and the closing of the speech made the division of labour explicit, calling for governments to provide the right policy background and the business community to do its part to invest[1].
The Number That Should Bother You
Of everything in the speech, one sentence carries the most weight and is the least quoted: "In fact, the level of productivity in Canada's business sector is more or less unchanged from where it was seven years ago"[5].
Sit with the implication. Seven years of technological change, seven years of software adoption, seven years of capital investment across the economy, and output per hour in the business sector had gone essentially nowhere. That is not a cyclical dip. It is a plateau.
The context was that Canadian labour productivity had eked out a small gain at the end of 2023, but only after six straight quarters in which productivity fell[2]. And the comparison with the United States sharpens it further: Rogers noted that back in 1984 the Canadian economy was producing 88 percent of the value generated by the US per hour worked[3], a figure she cited precisely because the gap has since widened.
Rogers also identified the specific disappointment. Given how nimble companies had been during the pandemic, the Bank expected productivity to improve coming out of it as firms found their footing and workers trained back up, and that happened in the US economy but did not happen in Canada[2].
What Happened Next
The natural question after any emergency declaration is whether it worked. The available evidence says no.
Writing in September 2025, roughly eighteen months after the speech, C.D. Howe Institute president William Robson and research officer Mawakina Bafale reported that things had gotten worse. Output and incomes per Canadian were on track to fall in 2025 for the third year running. Real, price-adjusted machinery and equipment investment by Canadian businesses dropped nine percent in the second quarter, leaving M&E investment twelve percent below its level of just a year earlier[6].
Their assessment is worth quoting for its bluntness: at a time when we need to equip our workers with the newest and best tools available, investment is collapsing[6]. And their verdict on the original warning: Carolyn Rogers was right, weak business investment in Canada is a national emergency, and our workers are not getting the tools they need to compete and prosper[6].
The authors acknowledge that policy uncertainty and fears of economic weakness are affecting other countries too, while noting that is little comfort given the divergent US story[6]. They also note a genuine Canadian strength that is usually omitted from the doom framing: Canada has an edge in non-residential construction, thanks largely to the relative size of our resource industries, while the United States is ahead on other categories[6]. The problem is specific, not general.
What Productivity Is Not
The word does more damage than good in public discussion, because business audiences reliably hear it as an instruction to work harder, and that is close to the opposite of what the data says.
Rogers defined it plainly: labour productivity measures how much an economy produces per hour of work, and increasing productivity means finding ways for people to create more value during the time they're at work[5]. The denominator is hours. Adding hours does not improve the ratio; it can worsen it.
One commentator made the point with a useful comparison, observing that Canadians work only about 19 hours less per year than Americans[5]. Whatever the precise figure, the direction is the point: the Canada-US productivity gap is not explained by Canadians working less. It is explained by output per hour, which is a function of the tools, processes, skills and market conditions people work with, not of their effort.
For an owner, this reframing matters practically. If your response to a productivity conversation is to push the team harder, you have misread the problem and will likely make retention worse while moving the actual metric very little.
The Three Components
Productivity decomposes into three drivers, and each implies a different intervention[2]. Capital deepening concerns the tools and equipment available per worker. Labour composition concerns improving workers' skills and training. Multifactor productivity concerns using capital and labour more efficiently, which is essentially process and organizational design.
Rogers dialled in on two themes: weak business investment in machinery, equipment and intellectual property; and problems in the labour market matching skilled workers with the right jobs[3]. Those map to capital deepening and labour composition respectively.
The inclusion of intellectual property alongside machinery and equipment deserves emphasis, because it is where a services business finds itself in the diagnosis. A professional services firm or software company does not buy much machinery, but it does or does not invest in software, systems, process documentation and proprietary methods, and that is the intangible capital half of the same problem. Statistics Canada research cited in the speech's own references addresses exactly this: Wulong Gu's paper "Investment Slowdown in Canada after the Mid-2000s: The Role of Competition and Intangibles"[1].
The Small Business Finding, Handled Fairly
Rogers made an observation that lands awkwardly for the readership of a publication like this one, and it should be reported rather than skipped: she noted that Canada has a larger proportion of small businesses, which tend to invest less[3].
Two things are true about this and both matter. The compositional claim is a statistical observation about the economy's structure, not a moral judgment about small business owners. An economy weighted toward smaller firms will show lower average investment per worker even if every individual firm is behaving rationally, because scale genuinely affects the economics of capital investment. A machine that pays back over 40,000 units a year does not pay back over 4,000.
But the second truth is the one worth acting on. The fact that small firms invest less on average does not mean your firm's current investment level is optimal for your firm. The compositional explanation and the individual optimization question are separate, and an owner can accept the first while still discovering that the second has a different answer than they assumed. That is the specific value of running the analysis rather than accepting the aggregate as fate.
Competition, And The Interprovincial Link
Beyond investment and skills, Rogers pointed to weak competition in Canada, regulatory uncertainty, and what she called a risk-averse business culture[3].
On competition specifically, she said: "Canada's economy features many sectors where companies face limited levels of competition, whether from firms in other provinces, foreign rivals or new entrants," while allowing that every country has certain sectors it champions and there can be valid reasons to protect local businesses[3].
The phrase "from firms in other provinces" connects this file directly to Canada's internal trade reform, examined elsewhere in this publication. If limited interprovincial competition is a contributor to weak productivity, then the Canadian Mutual Recognition Agreement and the associated provincial legislation are, whatever their limits, aimed at a genuine mechanism rather than a symbolic one. It also means the productivity argument for internal trade liberalization is stronger than the direct trade-volume argument, which some economists consider overstated.
The competition point has an uncomfortable edge for incumbent business owners. Weak competition is comfortable if you are the incumbent, and the productivity case for more of it is a case against a position many established Canadian firms occupy and would prefer to keep.
The Risk-Aversion Claim
The risk-averse business culture point is the softest item in the diagnosis and deserves the most scepticism, because cultural explanations are unfalsifiable in a way the investment data is not.
That said, there is a specific and testable version of it. If Canadian firms systematically apply higher hurdle rates, demand shorter payback periods, or weight downside scenarios more heavily than comparable US firms when evaluating the same investment, that is a measurable behavioural difference with real consequences for capital deepening, and it connects to the loss aversion and escalation dynamics examined elsewhere in this publication. An owner can check their own version of this directly: what payback period does your business require before approving equipment or software, and where did that threshold come from? In our experience the answer is frequently a number inherited from a predecessor with no analytical basis, applied uniformly regardless of the risk profile of the specific investment.
One commentator offered a related structural argument about the pandemic period specifically: that COVID crisis aid packages to companies and laid-off individuals contained no productivity-related incentives, with part of the business loans not requiring repayment, so the incentives to be more productive were simply not there[5]. This is that commentator's interpretation rather than a Bank of Canada finding, and it is contestable, since emergency support was designed for survival rather than productivity. It is nonetheless a coherent hypothesis for why the expected post-pandemic productivity rebound materialized in the US and not in Canada.
The Skills Matching Problem
The third element of Rogers's diagnosis was a failure to properly integrate skilled immigrants into the Canadian workforce[3], framed more broadly as problems in the labour market matching skilled workers with the right jobs[3].
This is a policy problem in significant part, involving credential recognition regimes that sit largely with provincial regulators. But there is a firm-level component an owner controls. A business that screens for Canadian experience, or that treats foreign credentials as unverifiable rather than investing modest effort in assessing them, is participating in the matching failure at its own expense, since the cost is borne partly by the firm in the form of a smaller and more expensive candidate pool.
The labour mobility provisions in recent internal trade legislation address the interprovincial dimension of matching, though as discussed elsewhere in this publication those provisions reach federally regulated occupations directly and provincial ones only through separate provincial instruments.
The Measurement Gap Is The Opening
Buried in the coverage of the speech is the single most actionable finding in this entire file, and it is stated almost in passing: few employers actually measure productivity[2].
This reframes the whole problem for an individual business. A national statistic is not something an owner can move. A metric their own business does not currently track is something they can start tracking on Monday, and it is difficult to improve a variable nobody observes.
The measurement need not be sophisticated. Revenue per employee, or better, gross profit per full-time-equivalent hour, computed monthly and trended over three years, is enough to reveal whether a business is getting more output from the same input over time. Most owners have never seen that line for their own business. Many are surprised by it, and the surprise usually runs in one of two directions: either the business has been adding headcount to sustain output, which shows as a declining trend, or it has quietly improved and nobody noticed because revenue growth masked the composition.
The second-order value is comparative. Once the series exists, capital investment decisions can be evaluated against it: did the machine, the software, the process change actually move output per hour, and by how much, and over what period? That converts capital deepening from an article of faith into a testable proposition, which is precisely the discipline the aggregate data suggests Canadian business has been missing.
Why A Central Bank Cares At All
It is worth understanding why the Bank of Canada, whose mandate is price stability, spent a major speech on this, because the reasoning has a direct implication for business owners.
Rogers's argument was that increasing productivity is a way to protect the economy from future bouts of inflation without having to rely so much on the cure of higher interest rates[1]. The mechanism is straightforward: if output per hour rises, wages can rise without unit labour costs rising, which means pay increases need not translate into price increases. If productivity is flat, wage growth feeds more directly into inflation, and the Bank's only tool for containing that is rates.
The implication for a business owner is uncomfortable but clarifying. A country with flat productivity and rising wages is a country where the central bank will keep reaching for higher rates. The cost of capital an owner complains about and the productivity performance they may not be measuring are connected, and the connection runs through exactly the investment decisions the aggregate data says Canadian firms are not making.
Rogers framed the Bank's own contribution as providing the stability most conducive to risk taking and investment[1], which is an honest acknowledgment that monetary policy can create conditions for investment but cannot make the investment happen.
A Worked Case: The Machine Nobody Bought
A Canadian manufacturer had evaluated a piece of automation equipment three times over five years and declined it each time. The illustrative reconstruction below reflects a recurring pattern rather than a specific engagement.
Each rejection was individually defensible. The first time, the payback period exceeded the firm's informal three-year threshold. The second, a softening order book made the volume assumptions look optimistic. The third, interest rates had risen and the financing cost had grown.
What the firm had never done was compute its own gross profit per production hour and trend it. When it finally did, the series showed a slow decline across the same five years: output per hour had fallen while headcount and wages rose, and the margin compression the owner had attributed to input costs was substantially a productivity effect. The equipment, evaluated against that series rather than against a payback threshold of unknown provenance, looked materially different, because the counterfactual was not "keep the status quo" but "continue a documented decline."
The generalizable point is not that the firm should have bought the machine. It is that for five years it made a capital allocation decision without the one measurement that would have told it what declining to invest was actually costing. That is the micro version of the national picture Rogers described, and it is the version an owner can fix without waiting for policy.
What An Owner Can Actually Do
Start measuring, this quarter. Gross profit per FTE hour, monthly, trended back three years if the data exists. This is the step almost nobody takes and everything else depends on.
Interrogate your hurdle rate. Identify the payback period or return threshold your business applies to equipment and software, and establish where it came from. An inherited three-year rule applied uniformly is a decision-making artifact, not an analysis.
Count intangible investment as investment. Software, systems, process documentation and proprietary methods are capital in the sense that matters here, and a services business that concludes the productivity file does not apply to it because it buys no machinery has misread the diagnosis.
Test the automation counterfactual honestly. Evaluate an investment against your actual measured trend, not against an assumption that declining to invest holds performance constant.
Examine your own hiring screens. Credential and Canadian-experience screening participates in the matching failure at the firm's own cost.
Reassess your market boundaries. If limited interprovincial competition has protected your position, it has also likely dulled it, and the internal trade reforms discussed elsewhere in this publication cut both ways.
The Honest Caveat On Individual Action
A publication that ended here would be overselling, so the limit deserves stating plainly.
Nothing an individual Canadian business does will move national productivity statistics. The drivers Rogers identified include competition policy, immigration and credential recognition, regulatory certainty and interprovincial barriers, none of which sit with an owner. C.D. Howe directed its own recommendation at the federal budget rather than at business[6], and that is the right address for most of the file.
The case for individual action is narrower and, we think, sound: the same underinvestment that shows up as a national statistic shows up in an individual firm as compressed margins, rising unit labour costs, and a business that is harder to sell. An owner should act on capital deepening and measurement because it is in their own interest at the level of their own P&L, not because it will help the aggregate. If it happens to help the aggregate, that is a byproduct rather than a motivation, and framing it otherwise asks business owners to act against their own interest for a public good, which is not a strategy that has ever worked at scale.
The Limits Of This Analysis
Several caveats matter. The Rogers speech dates to March 2024 and the C.D. Howe follow-up to September 2025; more recent quarterly data was not obtained for this article, and readers should check current Statistics Canada productivity and investment figures rather than assume the position described here is current. Productivity measurement is genuinely contested among economists, particularly around how intangibles and quality improvements are captured, and the aggregate statistics carry real methodological uncertainty. The risk-averse culture element of the diagnosis is the least empirically grounded and is presented as such. The commentary about pandemic aid lacking productivity incentives is one commentator's interpretation and is contestable. The worked case uses a constructed illustration. Finally, this is macroeconomic analysis translated for operators, not economic research, and readers wanting the underlying evidence should consult the sources cited in the Bank's own speech, including Haun and Sargent on decomposing Canada's post-2000 productivity performance and Gu on the role of competition and intangibles in the investment slowdown.
Frequently Asked Questions
What exactly did the Bank of Canada say?
Has anything improved since?
Does this mean Canadians need to work harder?
Rogers blamed small businesses. Is that fair?
What is the single most useful step for a business owner?
Why does a central bank care about productivity?
References
- Rogers, C. (2024, March 26). Time To Break The Glass: Fixing Canada's Productivity Problem. Remarks to the Halifax Partnership. Bank of Canada, including references to Haun & Sargent (2023) and Gu (2024). bankofcanada.ca/2024/03/time-to-break-the-glass-fixing-canadas-productivity-problem
- Canadian HR Reporter. (2024, March 27). 'In Emergency, Break Glass': Bank Of Canada Leader Cites Problem Of Canada's Productivity, including the three productivity components and the observation that few employers measure productivity. hrreporter.com/news/hr-news/in-emergency-break-glass-bank-of-canada-leader
- The Globe and Mail. (2024, March 28). Bank Of Canada Warns Of Low Productivity 'Emergency,' Making It Harder To Control Inflation. theglobeandmail.com/business/article-bank-of-canada-warns-of-low-productivity-emergency
- Bank for International Settlements. (2024, March 26). Carolyn Rogers: Time To Break The Glass, Fixing Canada's Productivity Problem, central bank speech reproduction. bis.org/review/r240326e.htm
- Blue Ocean CMC. (2024, April 2). Bank Of Canada Senior Deputy Governor Carolyn Rogers: Time To Break The Glass, quoting the speech and offering the commentator's own observations on hours worked and pandemic aid incentives. blueoceancmc.com/post/bank-of-canada-carolyn-rogers-break-the-glass
- Robson, W., & Bafale, M. (2025, September 26). Time To Break The Glass On Canada's Investment Emergency. C.D. Howe Institute. cdhowe.org/publication/time-to-break-the-glass-on-canadas-investment-emergency
- Toronto Star / PressReader. (2024, March 27). Productivity In Need Of Boost, BoC Says. pressreader.com/canada/toronto-star/20240327/281964612722454
This article discusses central bank remarks, think-tank analysis and macroeconomic data, and is provided for general informational purposes. It is not investment, economic or business advice. Productivity data is revised and methodologically contested; consult current Statistics Canada figures for the present position.