The role of EU cohesion funds in Romanian labour productivity: Insights from machine learning and econometric modelling

Abstract

This study examines the impact of European structural and investment funds (ESIF) on labour productivity across Romania’s NUTS 2 regions from 2007 until 2020. This research aims to assess how ESIF investments influence productivity imbalance while also identifying key regional determinants of economic performance, including socioeconomic structure, institutional quality, and educational attainment. Utilising a hybrid methodology integrating machine learning for variable selection and econometric modelling for effect estimation, the analysis leverages Least Absolute Shrinkage and Selection Operator to pinpoint the most influential factors and fixed effects panel regression models to quantify regional impacts. The results reveal significant disparities in the effectiveness of ESIF investments across regions, with factors such as governance quality and initial levels of development moderating the productivity outcomes. Findings signify that regions with stronger institutional infrastructures and higher baseline productivity levels benefit more meaningfully from ESIF investments. This research advances the understanding of ESIF’s role in promoting equitable regional development and demonstrates the utility of combining machine learning and econometric techniques in policy evaluation.

Keywords

• regional labour productivity
• european structural and investment funds
• cohesion policy impact
• econometrics
• machine learning
• eu funds efficiency