AI Analysis: Employment insurance disqualifications and disentitlements, monthly, unadjusted for seasonality

Executive Summary

Statistics Canada's Table 14100004 provides over 83 years of monthly Employment Insurance disqualification and disentitlement data (January 1943 to February 2026), capturing 122,154 records across 16 regions and 10 categories with an average of 67,714 persons affected per month at the national level. The data is highly right-skewed — the median of just 129 persons sits far below the mean of 2,198 — driven by dramatic historical spikes, including an all-time peak of 171,882 persons in June 1976 and a modern anomaly of 101,270 persons in August 2022. Long-term trends reveal major cycles tied to economic recessions and policy shifts, with recent years (2020 onward) averaging 77,899 persons per month, above the historical norm.

Key Findings

• The dataset spans 998 unique months from January 1943 to February 2026, making it one of the longest-running labour statistics series in Canada, with 122,154 records across 16 regions and 10 disqualification/disentitlement categories.
• The all-time peak of 171,882 persons occurred in June 1976, while the program's earliest recorded low was just 254 persons in October 1943, reflecting its limited initial scope during World War II.
• The distribution is strongly right-skewed, with a mean of 2,198 persons nearly 17 times higher than the median of 129, and approximately 18.1% of all records recording a value of exactly zero.
• Three statistically significant anomalies were identified using rolling z-scores: October 1945 (z=2.81), August 2022 (z=2.80, 101,270 persons), and August 1982 (z=2.68, 159,180 persons), each linked to major post-war, pandemic-era, or recessionary disruptions.
• Post-2020 monthly averages of 77,899 persons exceed the long-term national monthly average of 67,714, suggesting elevated EI disqualification activity likely influenced by pandemic-era policy changes.
• The middle 50% of all values fall between just 10 and 925 persons (IQR), underscoring that while most monthly regional counts are modest, a small number of high-volume observations drive the overall statistics.
• The 160 unique time series vectors (16 geographies × 10 disqualification types) enable granular regional and categorical comparisons, with Voluntary Quit and Misconduct historically representing the highest-volume individual disqualification categories.