Exploring Spatial Analysis Methods for Socio-Economic Profiling in the European Union Nuts2 Level

Socioeconomic profiling remains an ongoing priority for both researchers and public institutions, especially in the highly volatile context of the European Union. Therefore, in the current paper, we aim to compute several spatial analysis methods to extract observations regarding the performance of different development axes, at the NUTS2 regional level. For that objective, we will define the socioeconomic profile through five dimensions: economy, labour, science and technology, demographics, and education, each of them being represented by a specific indicator: GDP per capita, the unemployment rate, GERD, the median age of the population, and the participation rate in education. We consider this selection to be adequate for constructing a robust perspective, that can encapsulate the main characteristics and dynamics at the regional level, while also remaining concise. In terms of methodology, we will employ three methods, starting from Moran's I test for spatial autocorrelation, and then proceeding to spatial clustering through the K -Means algorithm. As a last step, we will also compare a linear regression model to a geographically weighted one, to decipher whether the spatial factor plays a role in defining the relationship between variables. Finally, throughout the application, our focus will be on interpreting the results and identifying local specificities, as well as differentiation points, which can then be integrated into the broad endeavour of constructing a socio-economic profile of the European Union regions.