Research

Canada faces persistent skilled labour shortages, yet only about 20% of apprenticeship registrants complete within the expected program duration. Apprenticeships are the primary pathway into the skilled trades, combining paid on-the-job training with technical instruction over a 3–4 year program. To encourage completion, the federal government introduced level-completion and one-time completion grants beginning in 2007. The existing grant structure was discontinued in March 2025, with proposals for a substantially more generous replacement currently under consideration, making counterfactual analysis of grant design particularly timely.

I build a dynamic discrete choice model of occupational choice and training investment estimated on linked administrative Canadian data from 2008 to 2020. I model individuals' career decisions as forward-looking choices that trade off short-run wage penalties during apprenticeship training against long-run completion premiums, with sector-specific human capital accumulating over time. The model allows prior work experience to affect program outcomes, reflecting the prevalence of delayed entry into Canadian apprenticeship programs. The estimated model is used to simulate counterfactual grant policies and their effects on enrolment, completion, and career trajectories.

This paper estimates separately the early career returns to apprenticeship participation and completion in the skilled trades sector, using rich longitudinal data from British Columbia spanning 1992–2022. This includes records on the universe of high school graduates in 2008–2011, apprenticeship enrollees, university students, and their respective tax records. I estimate a treatment effects model that separately identifies the returns to program enrolment and completion. Using propensity score matching, individuals who choose apprenticeship paths are paired with similar individuals who chose different paths.

I find that the early-career return to apprenticeship enrolment is approximately 25% relative to never enrolling, compared to 12% for a four-year university degree. Apprenticeship dropouts earn more than non-enrollees, reflecting sorting into well-paying industries and firms rather than lost schooling investment. University dropouts, by contrast, earn less than non-enrollees, consistent with the opportunity cost of foregone work experience.

We explore the extent of mismatch between job seekers and vacancies in Quebec. Mismatch may occur when those seeking jobs are concentrated in segments of the labour market, e.g. industries, with low job finding probabilities. Job finding probabilities might be low in some segments of the labour market because employers open few positions or because it is difficult to match employers and employees in these segments. Mismatch reduces the job finding rate and increases unemployment. In Quebec, unemployment has recently begun to rise and reached 6.3% in June 2025. In such situations, when the labour market softens, the policy discourse tends to focus on mismatch as one possible explanation. Mismatch in Quebec has increased between 2016 and 2025. This increase is however small compared to the overall level of unemployment. Much of the increase in mismatch results from an increase in the share of vacancies posted in health and social services that is not matched by an equivalent increase in job seekers in this sector.