AI Analysis: Average weekly hours for employees paid by the hour, by industry, monthly, unadjusted for seasonality

Executive Summary

Statistics Canada's Table 14100255 tracks average weekly hours for hourly-paid employees across 423 NAICS industry categories from January 2001 to February 2026, revealing a remarkably stable national average of approximately 32.2 hours per week with only a modest 0.6-hour decline over 25 years. The most meaningful variation in working hours is driven by industry type rather than time, with goods-producing sectors consistently outpacing service-producing industries, and individual sector averages ranging dramatically across the economy. A small number of data anomalies — most notably a 514.8-hour outlier for independent artists in December 2025 — warrant attention before using this dataset for precision analysis.

Key Findings

• Average weekly hours across all 211,617 records is 32.2 hours (median 33.1 hours), with the middle 50% of observations falling between 28.4 and 36.2 hours and a standard deviation of 5.64 hours.
• Over the full 25-year period (2001–2026), average weekly hours declined only slightly from 31.3 to 30.7 hours — a change of just 0.6 hours — indicating long-term structural stability in Canadian hourly work patterns.
• Goods-producing industries consistently report higher average weekly hours than service-producing industries, representing the most significant categorical divide in the dataset.
• Year and month have near-zero correlations with weekly hours (r=0.020 and r=0.008 respectively), confirming that industry category — not time — is the primary driver of variation in average weekly hours.
• 181 data points were flagged as outliers using the 3×IQR rule, including an extreme high of 514.8 hours for 'Independent artists, writers and performers [7115]' in December 2025 and a low of 7.9 hours for 'Radio and television broadcasting [5151]' in January 2016, both likely anomalies.
• The dataset covers 423 unique NAICS industry classifications and 846 unique time-series vectors, with approximately 15.4% of values missing — typical of data suppression at finer sub-sector levels.
• Unadjusted seasonal fluctuations are preserved throughout the dataset, and a visible dip in the 2020 time-series is consistent with the economic disruption caused by the COVID-19 pandemic.