Spatial dynamics of ICT employment in Europe: SAR analysis of the role of population and Internet infrastructure

Abstract

Introduction. In the modern era of digital transformation, information and communication technologies (ICT) have become a key factor in economic development. The labor market in the ICT sector exhibits significant geographical heterogeneity, necessitating the study of spatial dependencies that influence ICT employment levels in different countries. The use of spatial econometric models allows for the assessment of regional interconnections and the identification of key factors affecting the concentration of ICT specialists.

Purpose. This study aims to analyze the impact of population size and internet consumption on ICT employment in European countries using a Spatial Autoregressive (SAR) model. Particular attention is given to assessing the significance of spatial interdependencies and their influence on the development of the digital economy.

Research methods. To achieve the research objective, spatial econometric methods were employed, specifically Spatial Autoregressive (SAR) modeling. The Maximum Likelihood Estimation (MLE) method was used to estimate model parameters. Spatial dependence was tested using statistical measures such as Moran’s I and Geary’s C, and additional checks were performed for multicollinearity and heteroskedasticity.Creating a spatial weight matrix using the Queen contiguity method to determine spatial relationships between objects, in particular countries, based on shared borders.

Results. The findings indicate that population size is the most significant factor determining ICT employment levels, while internet consumption has a smaller but still positive effect. The spatial lag was found to be insignificant, suggesting a weak influence of neighboring countries on ICT sector development. Residual analysis revealed that predicted values closely align with actual data, although some regional deviations exist due to unaccounted socioeconomic factors.

Prospects. Further research could focus on expanding the model by incorporating additional variables such as education levels, investment in technology parks, the number of startups, and government support for digital initiatives. Another promising direction is the application of panel spatial models to account for dynamic changes in the ICT industry.