Digitalisation and Productivity in Italian Enterprises: Patterns of ICT Adoption and Firm Performance

As a driver of competitiveness, digitalisation is central to Europe’s economic agenda, yet adoption patterns remain heterogeneous, particularly between large firms and Small and Medium-sized Enterprises (SMEs). This thesis explores the micro-level associations between Information and Communication Technology (ICT) adoption and productivity in Italian enterprises, measured through value added per employee, thereby addressing a gap in the literature, which is often limited to macro indicators or single-technology studies. The analysis integrates three official Istat 2022–2023 datasets — the ICT Survey, the Structural Business Statistics (SBS) Frame, and the Statistical Register of Enterprise Groups — into a unified cross-sectional framework. Methodologically, Multiple Correspondence Analysis (MCA) and Exploratory Factor Analysis (EFA) are applied. The results are then used in two Ordinary Least Squares (OLS) regression models to examine associations with productivity. Findings reveal complementarities in ICT adoption, particularly in the joint use of computational power and management cloud services, Customer Relationship Management (CRM) and Business Intelligence (BI), Enterprise Resource Planning (ERP) systems and electronic sales solutions, as well as data analytics and data-sharing practices. Italy’s digitalisation emerges as highly polarised: large, multinational, high-tech firms adopt more advanced tools and achieve higher productivity, while SMEs and traditional sectors lag behind. Regression results confirm that digital sophistication and decision-support technologies (MCA factors) are positively associated with performance, though infrastructural tools alone may contribute less if adopted in isolation. The second OLS model further highlights that the integrated use of ERP and BI software, websites for professional visibility and market reach, infrastructural cloud solutions, and customer data exploitation are positively linked to productivity. Artificial Intelligence (AI) adoption, still limited and often hindered by skill shortages and high costs, shows contrasting effects. However, self-learning AI used in Research and Development (R&D) demonstrates a positive relationship. Overall, this thesis contributes to understanding how digital technologies cluster and influence firm performance, offering policy and managerial implications focused on technological integration and SME digital diffusion. Future research should adopt a panel approach to assess causality and explore whether the timing of AI-driven productivity gains varies across industries and application domains.