Hiring and job separation rates in the era of generative artificial intelligence: Early evidence
A · Strong evidence A short Statistics Canada report by Tahsin Mehdi and Marc Frenette on 20 years of Labour Force Survey microdata, split by an established AI exposure index; official data and a stated method, though descriptive with no causal design.
Summary based on the full text·Original: about 8 minutes to read
In short
A Statistics Canada report released 28 October 2026 that looks at hiring and job separation rates, rather than employment levels, to find early signs of AI effects in Canada. Youth hiring has fallen over the past five years, but it fell most in jobs least exposed to AI, and in coding jobs it began falling in early 2022, before ChatGPT. Separation rates are stable. The authors find no broad evidence so far that AI has cut hiring or raised separations, even for young workers in jobs most open to substitution.
Key numbers
- 4.0% to 3.3%monthly hiring rate of workers aged 15 to 29 in occupations less exposed to AI, before ChatGPT versus mid-2026
- 3.2% to 3.0%same rate in highly exposed occupations more amenable to substitution
- 2.5% to 1.9%same rate in coding-intensive occupations
- 5.8%15-year high of youth hiring in coding-intensive jobs, in the six months to November 2021, before the decline began
- 2.2%youth job separation rate in highly exposed, highly complementary jobs, unchanged across the period
As reported by the source.
English·Web page·opens www150.statcan.gc.ca
What it reports
- Hiring rate, ages 15 to 29, six months to November 2022 versus six months to August 2026: high exposure and high complementarity (doctors, nurses, teachers, engineers) 2.9% to 2.5%; high exposure and low complementarity (coding jobs, financial analysts, sales reps, data entry) 3.2% to 3.0%; low exposure (trades, physical work) 4.0% to 3.3%.
- Coding-intensive occupations: 2.5% to 1.9%. The decline started in early 2022, after a 15-year high of 5.8% in the six months to November 2021.
- Hiring rates for ages 30 to 49 and 50 to 64 were lower and more stable.
- Youth separation rates: about 2.2% in high-exposure, high-complementarity jobs, about 2.9% in high-exposure, low-complementarity jobs, about 3.2% in low-exposure jobs, all within historical ranges. Coding jobs rose from 2.4% to 3.5%, but the series is volatile on a small sample.
- Weak post-pandemic youth employment growth found in earlier work seems driven mainly by slower hiring, continuing earlier trends, not faster separations.
How they got it
Labour Force Survey microdata for employees aged 15 to 64, January 2006 to August 2026. New hires are respondents with one month of tenure; separations are new hires minus the net monthly change in paid employment. Series are seasonally adjusted with X-13ARIMA-SEATS and smoothed with six-month moving averages. Occupations are grouped by the complementarity-adjusted AI occupational exposure index (Felten et al.; Pizzinelli et al.).
Limits
- Descriptive comparison of periods; population shifts, trade shocks and post-pandemic adjustment are named as possible drivers but not separated.
- Exposure is assigned by occupation from a US-built index.
- Small samples for subgroups such as coding jobs.
- Covers employees only, not the self-employed.
Why it matters for companies
Claims that AI is already eliminating entry-level hiring often rest on one country's data or on postings. Canada's official flow data show youth hiring falling everywhere, faster in jobs AI touches least, with coding jobs turning down before ChatGPT. For workforce planners this cautions against attributing a hiring slowdown to AI without checking the trend before 2022, and suggests watching hiring rates by exposure group, not headcount totals, as the early signal.
Summary based on the full text. Read the original for the full argument and data.Read the original on www150.statcan.gc.ca