The Contribution of Artificial Intelligence (AI) to Labor Productivity Growth : A technical frontier approach

Sammanfattning

This thesis investigates the extent to which Artificial Intelligence (AI) contributes to labor productivity growth within European Union (EU) member states. Motivated by a notable deceleration in Europe’s labor productivity since the 1970s and growing policy interest in AI’s transformative potential, the study employs a non-parametric Data Envelopment Analysis (DEA) to measure and decompose productivity changes from 2000 to 2020. Drawing on data from 22 EU countries, the analysis treats AI-stock (R&D expenditure per AI-related patent) as a distinct production factor alongside conventional inputs such as labor, human capital, capital formation, and non-AI patent-stock.The quantitative analysis results indicate that EU-wide labor productivity more than doubled during the two decades studied, yet cross-country variations are substantial. Across the sample, incorporating AI as a separate factor explains an additional 4% in average labor productivity growth beyond what is captured by traditional inputs alone. The decomposition reveals that AI acts as a complementary driver in conjunction with human capital accumulation and gross fixed capital formation, although its positive impact is uneven across member states. Overall, these findings underscore the relevance of AI-driven innovation in strengthening EU’s labor productivity performance. Policymakers and industry stakeholders might leverage these insights to design strategies that foster targeted AI investments and skill development, thereby reinforcing the region’s competitiveness and sustainable economic growth.