Why it’s tough to verify some of Mark Carney’s claims about ‘protected’ jobs

Prime Minister Carney unveils a new automotive strategy that includes investments to help the industry pivot amid tariffs and plans to protect autoworkers.

OTTAWA — Prime Minister Mark Carney has made a series of claims recently about how much the federal government’s support for tariff-stricken industries has protected jobs in Canada.

Some of his figures on the number of workers being supported come close to federal records. Experts say that while there are models that can help estimate job creation tied to federal programs, measuring their impact on the labour market is seldom an exact science.

The claim

Since March 2025, the federal government has announced multiple measures to protect Canadian industries and workers vulnerable to U.S. tariffs. These measures have included changes to make it easier to access employment insurance, large pools of funds to help companies hold on to their workers and policies that encourage domestic firms to buy Canadian.

Speaking at an auto parts manufacturer in Woodbridge, Ont. on Feb. 5, Carney touted the results of what he called “the most comprehensive set of trade resilience measures in Canada’s history.”

“Our measures have created and protected 18,000 jobs across steel, aluminum, lumber and the auto sector. They’ve prevented more than 20,000 layoffs,” he said.

“We provided income supports for more than 6,000 workers, with a total of 190,000 more expected to benefit, including in the auto sector.”

The facts

On the day Carney made that claim, The Canadian Press reached out to the Prime Minister’s Office to ask for the source of his figures.

The next day, Feb. 6, the PMO forwarded the request on to Employment and Social Development Canada.

After two deadline extensions, ESDC provided a short answer around 5 p.m. ET on Feb. 10, the following Tuesday.

That reply stated that the estimate of layoffs avoided — 20,000 — came from the federal government’s work-sharing program data, while the figure on workers receiving income support was taken from the employment insurance program.

The federal government’s work-sharing program offers income support for workers with reduced hours when their employer is facing a downturn in business outside the company’s control.

An ESDC web page tracking the program estimates that 18,621 layoffs had been prevented since March 2025 as of the week ending Feb. 7, 2026 — slightly below Carney’s figure. The program has approved 1,450 total agreements with employers representing 48,979 employees at a total cost of $307,818,851, according to the federal government’s statistics.

Another ESDC web page states that 8,360 people have become first-time recipients of employment insurance since April 1, 2025, though those figures are not divided by industry.

ESDC’s Feb. 6 response did not include Carney’s 18,000 jobs “created and protected” figure. The Canadian Press requested additional information related to those numbers the following day.

The department requested multiple extensions through the week and did not provide a response by the final deadline of noon on Feb. 13.

Tony Stillo is the director of Canadian economics at Oxford Economics and previously worked at the Ontario Ministry of Finance, where he modelled the labour market impacts of various government policies.

He said there are a number of ways to chart potential job impacts and some are more clear than others.

With direct financial supports such as the work-sharing program, Stillo said, applicants have to provide the federal government with regular and detailed payroll data that helps to inform statistics about the number of jobs affected, or the number that otherwise would have been lost.

“The jobs at risk is a bit of a subjective figure, but that’s kind of a reference point,” he said.

Stillo said economic models are fairly reliable when it comes to the number of jobs created, lost or maintained across a supply chain. A tiremaker might be able to hold on to more workers if the automotive company they supply is also doing more business, for example.

He said the models get less reliable when they examine “induced” job impacts — knock-on effects from a loss of income across the economy. If an autoworker isn’t stopping into his local dinner on the way into his shift, the person serving his coffee might see their job put at risk. Stillo said those second-order effects are harder to predict.

Stillo said governments typically don’t report induced job figures and default to more verifiable figures.

“That’s where I would draw the line, that induced effect. I would stick to the direct activity at the plant, employment in this case, and then the supply chain, not the re-spending of the incomes that have been supported,” he said.

Randall Bartlett, deputy chief economist at Desjardins, said it’s “standard fare” for governments to estimate the labour market effects of their policies. He also said it’s hard to know whether Carney’s 18,000-job figure is accurate without also knowing “what’s under the hood” of Ottawa’s models.

Outside of direct income supports, Bartlett said, the federal government will have to derive job estimates from comparisons to historic data.

Policies like Budget 2025’s proposed “productivity super deduction” — a measure allowing businesses to write off the full cost of investments like new equipment in year one — might be judged against how much similar policies drove business investment and job creation in previous years.

“Those are much harder to track and they’re really based on different estimation approaches and they’ll give different results,” Bartlett said.

Statistics Canada said in its January labour force survey that employment in tariff-sensitive manufacturing was down 51,000 positions from a year earlier.

Stillo said in a recent report that tracking the impact of the trade war on the jobs market has been challenging, in part because the monthly labour force survey has diverged at times this year from StatCan’s survey of employment, payrolls and hours — a separate measure of employment that’s usually less volatile but also less timely than the labour force survey.

Stillo said when tracking individual data sets is difficult, it’s more important to take a step back.

“We think the big picture is the economy is struggling to grow and will continue to do so because of the trade war, the uncertainty related to that,” he said.

Stillo added, however, that modestly stimulative interest rates from the Bank of Canada and fiscal policy supports from multiple levels of government are offering “tailwinds” to the economy.

While exact figures are hard to nail down, Bartlett said the labour market has proven surprisingly resilient to the trade war so far and the impact of the federal government’s policies has thus far been “positive.”

“We can quibble over individual numbers. Reasonable people can disagree on what those numbers are and what does constitute reasonable. I think they have helped to prevent certainly some layoffs, maybe added a little bit more to hiring than we would’ve seen otherwise,” he said.

Bartlett said Ottawa’s task now is to decide whether these programs are continuing to work as designed or need to evolve, and whether taxpayers are getting “bang for their buck.”

This report by The Canadian Press was first published Feb. 15, 2026.

Craig Lord, The Canadian Press