Is Canada's labour market weakening?
Employment rose over the year but fell in the latest month. Youth unemployment remained well above the prime-age rate; entry-level hiring outcomes remain unresolved.
Executive summary
The current record is mixed. Employment grew faster than the working-age population over the year, yet employment fell in the latest month. The annual ratio therefore does not establish stronger current-month hiring. Youth unemployment remains well above the prime-age rate, with a ratio close to its level a year earlier. The Starting Work Question adds NEET estimates, real employee wages by age and selected graduate-cohort earnings. Current first-job hiring, broader graduate outcomes and seniority remain unresolved. The cited results below give each measured comparison and its period.
Labour force survey population, age 15+ (SA): all-time high (34,854,900 in Sep 2026)
The latest value of labour force survey population, age 15+ (sa) (34,854,900, Sep 2026) is the highest in the series history since 2006 (243 months of data). A record marks where the series sits in its own history; it carries no causal claim.
Technical details
Youth unemployment ran 2.32x the prime-age rate in Sep 2026
Unemployment for ages 15-24 was 13.0% (seasonally adjusted) vs 5.6% for ages 25-54 in Sep 2026 — a ratio of 2.32, vs 2.43 a year earlier. Both rates are measured on the same seasonally-adjusted basis from the same monthly survey (StatsCan 14-10-0287), so the ratio is not distorted by basis mismatches the way cross-survey comparisons can be.
Technical details
Employment grew 0.52 employed people per person aged 15 and over added over the year to Sep 2026
In the 12 months to Sep 2026, employment (age 15+, seasonally adjusted) rose by 94,900 while the LFS working-age population estimate rose by 183,000 — 0.52 employed people added per additional working-age person, vs 0.37 a year earlier. A ratio at or above 1 suggests employment growth is keeping pace with (or exceeding) working-age population growth; below 1 suggests the labour market is absorbing the growing working-age population more slowly than it is growing. This says nothing about participation-rate changes, part-time/full-time mix, hours worked, or sector/regional composition. In the latest month alone, employment changed by -68,300 people. The annual comparison's base month changed by +53,300 people from its preceding month; moving the base therefore changes the annual employment gain by -121,600. A higher annual ratio does not by itself establish stronger current-month employment.
Technical details
What the record shows
Read the monthly and annual comparisons together. A change in the annual comparison’s base month can raise the annual employment gain even when the latest month loses employment. The absorption ratio is a headcount comparison; it does not adjust for participation, hours or job mix.
Labour force survey population, age 15+ (SA): all-time high (34,854,900 in Sep 2026)
The latest value of labour force survey population, age 15+ (sa) (34,854,900, Sep 2026) is the highest in the series history since 2006 (243 months of data). A record marks where the series sits in its own history; it carries no causal claim.
Technical details
Employment grew 0.52 employed people per person aged 15 and over added over the year to Sep 2026
In the 12 months to Sep 2026, employment (age 15+, seasonally adjusted) rose by 94,900 while the LFS working-age population estimate rose by 183,000 — 0.52 employed people added per additional working-age person, vs 0.37 a year earlier. A ratio at or above 1 suggests employment growth is keeping pace with (or exceeding) working-age population growth; below 1 suggests the labour market is absorbing the growing working-age population more slowly than it is growing. This says nothing about participation-rate changes, part-time/full-time mix, hours worked, or sector/regional composition. In the latest month alone, employment changed by -68,300 people. The annual comparison's base month changed by +53,300 people from its preceding month; moving the base therefore changes the annual employment gain by -121,600. A higher annual ratio does not by itself establish stronger current-month employment.
Technical details
The youth gap
The youth unemployment rate remains well above the prime-age rate. The population/youth timing and co-movement results below use quarterly windows that end earlier than the latest monthly unemployment comparison. They describe selected historical patterns and cannot identify a cause or establish outcomes for people entering work.
Youth unemployment ran 2.32x the prime-age rate in Sep 2026
Unemployment for ages 15-24 was 13.0% (seasonally adjusted) vs 5.6% for ages 25-54 in Sep 2026 — a ratio of 2.32, vs 2.43 a year earlier. Both rates are measured on the same seasonally-adjusted basis from the same monthly survey (StatsCan 14-10-0287), so the ratio is not distorted by basis mismatches the way cross-survey comparisons can be.
Technical details
Population growth, % y/y and Unemployment rate, age 15-24 (SA) moved in opposite directions over the 12 quarters to 2026-07 (r=-0.52). Correlation is not causation.
Movements in Unemployment rate, age 15-24 (SA) have tended to precede movements in Population growth, % y/y by about 4 quarters in this window (lagged r=-0.96 vs contemporaneous r=-0.56). This is a timing pattern only — a candidate relationship, not a causal claim.
Cross-currents in the data
The refreshed calculation still does not produce the previously featured unemployment–vacancy lead-lag result. That absence does not prove there is no relationship; it means the current data do not meet that detector’s rule. The remaining results below have their own windows and thresholds. None establishes causation or a hiring forecast. The September update also no longer meets the rule for the formerly featured co-movement between the immigrant unemployment gap and national unemployment. No weaker threshold or causal inference replaces it.
Employment, age 15+ (SA) and Population growth, % y/y moved in opposite directions over the 12 quarters to 2026-07 (r=-0.94). Correlation is not causation.
Movements in Employment, age 15+ (SA) have tended to precede movements in Population growth, % y/y by about 4 quarters in this window (lagged r=-0.95 vs contemporaneous r=-0.69). This is a timing pattern only — a candidate relationship, not a causal claim.
The detected level shift in Unemployment rate (SA) (around 2024-03) occurred in the same month as the detected shift in Job vacancy rate (around 2024-03). Concurrence is a timing fact and does not attribute causality.
The detected level shift in Job vacancy rate (around 2024-03) occurred in the same month as the detected shift in Unemployment rate, age 15-24 (SA) (around 2024-03). Concurrence is a timing fact and does not attribute causality.
One commitment, no way to check it yet
A government jobs commitment is on the public record. No official series currently attributes job creation to the mechanism the commitment names, so no pace verdict can be published for it.
Create 250,000 new AI-related jobs over five years, as part of an additional $200 billion of economic growth
Issuer: Government of Canada (Prime Minister's Office, 'AI for All' national AI strategy). Stated in AI for All strategy launch.
Latest observed: — against target 250,000.
What's coming
One dated, cited change to labour-market support is already on the record and upcoming.
Temporary EI relief measures expire
Temporary Employment Insurance measures — waiver of the one-week waiting period, suspension of the treatment of monies paid on separation, and an extra 20 weeks of income support for long-tenured workers — were extended by regulation to October 10, 2026. After this date, standard EI rules apply unless further extended.
What the record cannot say
The record distinguishes resolved source-availability gaps from remaining gaps in entry-level hiring and graduate outcomes. NEET estimates, real wages by age and selected graduate-cohort earnings are now available in the Starting Work Question. They do not identify current first-job vacancies or follow every person into stable work.
Read the Starting Work comparisons →
Is entry-level and early-career labour demand actually shrinking, or is this a composition effect?
What is missing: No official Statistics Canada series isolates entry-level or early-career job demand. The Job Vacancy and Wage Survey (table 14-10-0432 and related) measures vacancies by industry, province, and occupation, but not by required experience level or candidate seniority. The Bank of Canada has itself named 'entry-level job vacancy declines' as something it watches without citing a dedicated official series for it.
Consequence: Countability publishes the youth-vs-prime-age unemployment ratio and the overall vacancy rate as the closest available proxies; neither directly measures entry-level demand, and this is stated wherever they are shown together.
Is AI adoption creating, displacing, or reshaping jobs — and is the government's 250,000-AI-job target on track?
What is missing: No official series attributes job creation, displacement, or reshaping to AI adoption specifically. Statistics Canada's Labour Force Survey and business-adoption surveys measure aggregate employment and technology use separately; there is no published methodology linking the two, and the government has not stated how the 'AI for All' strategy's 250,000-job target will be measured.
Consequence: This commitment is published as a registered commitment without a compliance claim. Countability will not publish INFERRED claims attributing employment changes to AI adoption absent an official attribution methodology.
How are recent graduates actually faring in the labour market?
What is missing: Administrative earnings are now qualified from Statistics Canada table 37-10-0301-01: selected Canadian graduate cohorts through 2022, observed two years later. These data exclude some graduates and do not measure contemporaneous hiring, underemployment or whether jobs match fields of study.
Consequence: Countability publishes the available cohort earnings, with their lag and exclusions. It does not use the youth age group as a substitute for recent graduates or treat earnings among selected earners as an employment rate.
How many Temporary Foreign Worker positions are employers being approved to fill (LMIA volumes)?
What is missing: A real ESDC/IRCC open-data LMIA employer list exists, but is published as separate per-quarter CSV resources rather than one continuous machine-readable series, and carries the publisher's own completeness caveat: it 'excludes all personal names' and 'is not complete and does not reflect all employers who requested or received an LMIA.' It counts positions on positive LMIAs, including subsequently cancelled ones, not a clean approval-volume metric.
Consequence: Countability does not read LMIA volumes for this reading. Work-permit flow series (already published, IRCC IMP+TFWP monthly) are shown as related context instead, with the concepts kept explicitly distinct (permits issued vs. LMIA positions approved are different steps in the process).
What has since been established
How many young Canadians are not in employment, education, or training (NEET)?
What is missing: The earlier source-absence finding was corrected on October 5, 2026. Statistics Canada table 37-10-0196-01 supplies national first-quarter NEET averages through 2025, with age groups and standard errors.
Consequence: Countability now publishes this bounded NEET comparison. It measures a first-quarter average among youth aged fifteen to twenty-nine in the provinces, not continuous year-long exclusion from work or education.
Resolved 5 October 2026: Source-availability assertion corrected by direct qualification of Statistics Canada 37-10-0196-01 (first-quarter NEET averages through 2025). See the starting-work Big Question; remaining scope limits are explicit.
Are youth wages keeping pace with inflation?
What is missing: The earlier source-absence finding was corrected on October 5, 2026. Statistics Canada table 14-10-0064-01 supplies annual median hourly wages by age through 2025; matched annual all-items CPI supplies the deflator.
Consequence: Countability now compares nominal and inflation-adjusted hourly wages of employees aged fifteen to twenty-four and twenty-five to fifty-four. This is a comparison of people in employment, not the wage offers received by all job seekers or a panel following the same workers.
Resolved 5 October 2026: Source-availability assertion corrected by direct qualification of Statistics Canada 14-10-0064-01 (annual wages by age through 2025) with matched all-items CPI. See the starting-work Big Question; remaining scope limits are explicit.
Source trail & method
- Labour force characteristics, monthly, seasonally adjusted — Statistics Canada
- Job vacancies, payroll employees, job vacancy rate, and average offered hourly wage, monthly, seasonally adjusted — Statistics Canada
Each claim carries its evidence tier, confidence and source. Correlation means observed co-movement, not cause. Missing links in the evidence are identified, and every figure belongs to a cited member of the published record.
Where this sits in the record
This Deep Dive and its underlying question are published in full. See Work with Countability for a commissioned one.
Work with Countability