AI Analysis: Employment insurance beneficiaries by type of income benefits, monthly, unadjusted for seasonality

Executive Summary

Statistics Canada's Table 14100009 tracks nearly 30 years of monthly Employment Insurance beneficiaries across 20 benefit types, revealing a strongly right-skewed distribution dominated by Regular Benefits and punctuated by a historic COVID-19 surge that drove claims to an all-time peak of 2,097,070 persons in January 2021. The dataset's long time horizon (January 1997–February 2026) captures multiple economic cycles, with the pandemic representing a 2.2x spike over the pre-COVID monthly average of ~793,000 beneficiaries. As of February 2026, beneficiary counts remain at approximately 1,017,180 — roughly double pre-pandemic levels — suggesting a lasting structural shift in EI usage.

Key Findings

• The dataset spans 350 monthly observations from January 1997 to February 2026, covering 20 distinct EI benefit types across 6,784 valid records.
• The all-time peak of 2,097,070 beneficiaries occurred in January 2021, representing a 2.2x surge over the pre-COVID monthly average of approximately 793,000 persons.
• The distribution is strongly right-skewed, with a mean of 133,404 persons nearly 9 times higher than the median of 15,135, reflecting a few very large benefit categories pulling the average up.
• September 2020 recorded the dataset's minimum of 345,420 beneficiaries, likely caused by the federal CERB program temporarily diverting claimants away from traditional EI.
• As of February 2026, monthly beneficiaries stand at 1,017,180 — approximately double pre-pandemic levels — indicating persistently elevated EI demand years after the initial COVID shock.
• Regular Benefits is the dominant category by average monthly volume, while specialized programs such as Compassionate Care and Family Caregiver Benefits represent smaller but growing segments.
• Correlation analysis across all 20 benefit types shows that many categories move together in response to shared economic shocks, yet rolling z-score analysis flagged zero formal statistical outliers, as the prolonged COVID surge gradually shifted the local baseline.