AI Analysis: Small area estimates of labour force characteristics for sub-provincial areas, monthly, unadjusted for seasonality

Executive Summary

Statistics Canada Table 14100480 provides monthly labour force estimates across 659 sub-provincial areas from November 2021 to June 2024, tracking Employment, Employment Rate, and Unemployment Rate with associated confidence intervals across 250,000 records. Employment figures are highly skewed — ranging from 10 persons to nearly 3.7 million — reflecting the vast size differences between sub-provincial areas, with Toronto CMA recording the highest average employment at 3,719,860 persons. The strongest analytical signal in the dataset is the strong negative correlation (r = -0.71) between Employment Rate and Unemployment Rate, confirming these measures reliably move in opposite directions across Canadian sub-provincial labour markets.

Key Findings

• The dataset covers 659 unique sub-provincial areas across Canada over 32 monthly periods, yielding 250,000 records that capture fine-grained regional labour market conditions.
• Employment is highly right-skewed, with a median of just 5,170 persons versus a mean of 32,558, and values ranging from 10 to nearly 3.7 million persons — driven by large urban centres like Toronto CMA at the top.
• Employment Rate and Unemployment Rate show a strong negative correlation (r = -0.71), making them the most informative pair for assessing sub-provincial labour market health, while total Employment count has only weak relationships with either rate measure.
• Aggregate employment across all sub-provincial areas declined 22.8% from approximately 19.4 million (November 2021) to 15.0 million (June 2024), though this trend may partly reflect seasonal patterns given the data is unadjusted for seasonality.
• Employment Rate is relatively stable and normally distributed with a mean and median both near 56% and a standard deviation of only 9 percentage points, while Unemployment Rate is right-skewed with a median of 6% but reaching as high as 43% in some areas.
• Outlier analysis identified 1,488 employment records outside extreme IQR bounds and flagged several geographic areas with mean values more than 3 standard deviations above the overall average, with no zero-value records found across 58,701 estimate records.
• Data is provided with full uncertainty quantification — each record includes the estimate, its standard error, and lower/upper 95% confidence interval bounds — enabling statistically rigorous comparisons across small areas.