Artificial Intelligence in Public Employment Services: Machine Learning, Job Match Quality, and Fairness in Labour Market Matching

This dissertation examines the use of artificial intelligence and machine learning in public employment services. Using comprehensive administrative labour market data, it investigates three key dimensions of algorithm-based occupational recommendations: predictive accuracy, job match quality, and fairness. The results demonstrate that machine learning methods outperform conventional statistical approaches in predicting occupational transitions. Furthermore, a matching index is developed that incorporates expected employment stability and wages alongside predicted employment probabilities. Finally, the dissertation examines potential biases and inequalities in algorithmic recommendations from an intersectional perspective. The findings highlight that the responsible use of artificial intelligence in public employment services requires considering not only predictive performance but also the quality and fairness of occupational recommendations.