Machine learning and the labour market: A portrait of occupational and worker inequities in Canada

Introduction Machine learning (ML) is increasingly used by Canadian workplaces. Concerningly, the impact of ML may be inequitable and disrupt social determinants of health. The aim of this study is to estimate the number of workers in occupations highly exposed to ML and describe differences in ML exposure represents according to occupational and worker sociodemographic factors.

Methods Canadian occupations were scored according to the extent to which they were made up of job tasks that could be performed by ML. Eight years of data from Canada’s Labour Force Survey were pooled and the number of Canadians in occupations with high or low exposed to machine learning were estimated. The relationship between gender, hourly wages, educational attainment and occupational job skills, experience and training requirements and ML exposure was examined using stratified logistic regression models.

Results Approximately, 1.9 million Canadians are working in occupations with high ML exposure and 744,250 workers were employed in occupations with low ML exposure. Women were more likely to be employed in occupations with high ML exposure than men. Workers with greater educational attainment and in occupations with higher wages and greater job skills requirements were more likely to experience high ML exposure. Women, especially those with less educational attainment and in jobs with greater job skills, training and experience requirements, were disproportionately exposed to ML.

Conclusion ML has the potential to widen inequities in the working population. Disadvantaged segments of the workforce may be most likely to be employed in occupations with high ML exposure. ML may have a gendered effect and disproportionately impact certain groups of women when compared to men. We provide a critical evidence base to develop strategic responses that ensure inclusion in a working world where ML is commonplace.