Publication: La productividad laboral en la era de la IA: Perspectivas a partir de Datos de Panel
Loading...
Files
Identifiers
Publication date
Reading date
Event date
Start date of the public exhibition period
End date of the public exhibition period
Authors
Yoshida, Hiroshi
Yenilmez, Meltem Ince
Advisors
Authors of photography
Person who provides the photography
Journal Title
Journal ISSN
Volume Title
Publisher
Universidad Pablo de Olavide
Abstract
Este estudio examina cómo la productividad laboral se ve afectada por los avances relacionados con la IA mediante el examen de un conjunto de datos de panel equilibrado de empresas que operan en tres áreas diferentes entre 2014 y 2023. El estudio utiliza modelos OLS combinados, de efectos fijos (FE) y de efectos aleatorios (RE) para evaluar los efectos de factores importantes como las patentes relacionadas y no relacionadas con la IA, la inversión en I+D, la mano de obra y la rotación de las empresas sobre la productividad laboral, la variable dependiente. Según los resultados, las patentes relacionadas con la IA tienen un notable impacto positivo en la productividad laboral. Por otra parte, la mano de obra tiene una correlación negativa con la productividad, lo que indica ineficiencias en la gestión de plantillas más grandes o rendimientos decrecientes a escala. Es interesante observar que la rotación de personal y la productividad están positivamente correlacionadas, lo que podría ser resultado de la optimización de la mano de obra o de la introducción de nuevas perspectivas y habilidades. En comparación con el modelo OLS agrupado, el modelo FE, que tiene en cuenta la heterogeneidad específica de las empresas, explica en torno al 45,7% de la varianza de la productividad. Las pruebas de diagnóstico verifican la resistencia de los modelos y su validez mejora con correcciones de autocorrelación y heteroscedasticidad.
This study examines how labour productivity is affected by AI-related developments by examining a balanced panel dataset of businesses that operate in three different areas between 2014 and 2023. The study uses pooled OLS, Fixed Effects (FE), and Random Effects (RE) models to assess the effects of important factors such as AI-related and non-AI patents, R&D investment, labour input, and company turnover on labour productivity, the dependent variable. According to the findings, AI-related patents have a notable positive impact on labour productivity, supporting earlier studies on technology-driven productivity improvements and highlighting the critical significance of AI innovation. Labour input, on the other hand, has a negative correlation with productivity, indicating either inefficiencies in managing larger workforces or diminishing returns to scale. It’s interesting to note that employee turnover and productivity are positively correlated, which could be a result of workforce optimisation or the introduction of new perspectives and abilities. In comparison to pooled OLS, the FE model, which takes firm-specific heterogeneity into account, explains around 45.7 % of the productivity variance. Diagnostic tests verify the models’ resilience, and their validity is improved by autocorrelation and heteroscedasticity corrections. These findings warn against inefficient labour use while highlighting the value of AI and R&D expenditures in boosting productivity. This study advances our knowledge of the dynamics of productivity, labour management, and innovation and guides company executives and policymakers as they navigate the AI-driven economy.
This study examines how labour productivity is affected by AI-related developments by examining a balanced panel dataset of businesses that operate in three different areas between 2014 and 2023. The study uses pooled OLS, Fixed Effects (FE), and Random Effects (RE) models to assess the effects of important factors such as AI-related and non-AI patents, R&D investment, labour input, and company turnover on labour productivity, the dependent variable. According to the findings, AI-related patents have a notable positive impact on labour productivity, supporting earlier studies on technology-driven productivity improvements and highlighting the critical significance of AI innovation. Labour input, on the other hand, has a negative correlation with productivity, indicating either inefficiencies in managing larger workforces or diminishing returns to scale. It’s interesting to note that employee turnover and productivity are positively correlated, which could be a result of workforce optimisation or the introduction of new perspectives and abilities. In comparison to pooled OLS, the FE model, which takes firm-specific heterogeneity into account, explains around 45.7 % of the productivity variance. Diagnostic tests verify the models’ resilience, and their validity is improved by autocorrelation and heteroscedasticity corrections. These findings warn against inefficient labour use while highlighting the value of AI and R&D expenditures in boosting productivity. This study advances our knowledge of the dynamics of productivity, labour management, and innovation and guides company executives and policymakers as they navigate the AI-driven economy.
Doctoral program
Related publication
Research projects
Description
Bibliographic reference
Revista de métodos cuantitativos para la economía y la empresa, ISSN-e 1886-516X, Vol. 40, 2025, págs. 1-18




