AI Analysis: Industrial capacity utilization rates, by industry

Executive Summary

Statistics Canada's Table 16100109 tracks quarterly industrial capacity utilization rates across 34 NAICS industry categories from 1987 to 2025, encompassing 5,224 data points with a long-run average of 79.4%. The dataset reveals a broadly symmetric distribution of utilization rates, significant variation across sectors (ranging from 42.1% to 108.4%), and a clear imprint of major economic disruptions on Canadian industrial output. The most recent reading of 78.5% (Q4 2025) sits below the historical average of 81.6%, suggesting modest slack remains in Canada's industrial base.

Key Findings

• The long-run average capacity utilization rate across all industries and periods is 79.4%, with a nearly identical median of 79.7%, indicating a symmetric distribution with moderate variability (standard deviation of 7.1%).
• Utilization rates span a wide range of 66.3 percentage points — from a record low of 42.1% to a record high of 108.4% — reflecting substantial differences in demand conditions and structural capacity across industries.
• The all-time peak for Total Industrial utilization was 86.4% in April 1988, while the historic trough of 69.6% occurred in April 2020, directly coinciding with COVID-19 pandemic shutdowns.
• The most recent quarterly reading (Q4 2025) stands at 78.5%, approximately 3 percentage points below the historical average of 81.6%, indicating that Canadian industry is currently operating with room to expand.
• The middle 50% of all observations fall within a relatively tight band of 74.9% to 84.1% (an IQR of 9.2%), suggesting that while extremes exist, most industries operate within a consistent range over time.
• Correlation analysis across the 34 industry categories reveals that many manufacturing sub-sectors move together through economic cycles, while certain sectors exhibit low or negative correlations, reflecting independent demand and supply drivers.
• Outlier detection using the IQR method identified statistically unusual utilization readings concentrated around major economic disruptions, with high-volatility industries accounting for the greatest share of anomalous data points across the nearly 40-year dataset.