AI Analysis: Usual hours worked by job type (main or all jobs), monthly, unadjusted for seasonality

Executive Summary

Statistics Canada Table 14100030 provides 50 years of monthly data (1976–2026) on usual hours worked by Canadians, spanning 234,570 records across 11 regions, 2 job types, 3 gender categories, and 9 age groups. The dataset is heavily right-skewed — with a median of 30.3 versus a mean of 2,158 — reflecting the mix of population counts (persons in thousands) and hours-based measures within the same table. Key structural patterns include a persistent gender gap in hours worked, strong cross-category correlations driven by population size, and identifiable anomalies tied to major economic disruptions including the COVID-19 pandemic.

Key Findings

• The dataset contains 234,570 records spanning January 1976 to February 2026, organized across 5,940 unique time series vectors and 11 geographic regions including a national total and all provinces.
• The distribution is strongly right-skewed, with a median value of 30.3 far below the mean of 2,157.98 and a standard deviation of 15,865.75, driven by the coexistence of person-count records (up to 11,199.3 thousand) and hours-based records (up to 445,026.5 hours) in the same dataset.
• Total employment in Canada (all genders, 15+) grew by approximately 20.8% between January 1976 and July 1979, with the full 50-year series ranging from a low of ~9,271K to a high of ~11,199K persons employed.
• A notable gender gap exists: Men+ report a higher mean of 2,220.60 and median of 28.8 hours compared to Women+ with a mean of 1,063.41 and median of 21.7 hours, reflecting structural differences in usual hours worked.
• Outlier detection using the 3×IQR method (bounds: -301.1 to 413.6) flagged 34,002 records — approximately 13.6% of the dataset — as extreme values, with anomalous months in the time series likely corresponding to recessions and the COVID-19 pandemic.
• Most 'Usual Hours Worked' categories are strongly positively correlated with one another across gender and age group breakdowns, a pattern largely driven by population size effects rather than independent behavioural variation.
• The dataset captures two units of measure — 'Persons in thousands' (184,754 records, median 14.5) and 'Hours' (49,816 records, median 47.3) — which must be interpreted separately to avoid misleading comparisons across categories.