Social exclusion represents a popular topic in the policy agendas of European governments, especially after the Great Recession. The existing literature highlights the presence of spatial patterns, although previous contributions consist of local or national level studies, lacking a broader continental perspective. This work resorts to regional data covering 22 EU countries and aims to characterise the nature of spatial patterns, controlling for socio-economic covariates. Using the Spatial Markov Chain Matrix, we find that the strong clusterisation process unfolded by previous studies tends to become less intense if socio-economic covariates are taken into account. Socio-economic factors represent in other words a containment cage that reduces the extent of neighbour influence.