Title: Dynamics of spatial autocorrelation: A new space-time state of mind
Authors: Solmaria Halleck-Vega - Wageningen University (Netherlands)
Roberto Patuelli - University of Bologna (Italy) [presenting]
Abstract: The dynamics of spatial autocorrelation, which is present in most spatial data, are explored. Usually it is accounted for either in the error term and/or using other spatial econometric and statistic techniques. It can thus either be treated as a nuisance or substantive phenomenon. In contrast to studies focusing on including spatial effects in regression models, we explore actual changes in spatial autocorrelation over time. This offers a distinctive perspective. As a first step, it is useful to appraise the dynamics and relationships between time series data and the respective evolution of spatial autocorrelation. Then, to better understand whether trend-cycle and/or seasonal components have a role to play in explaining spatial autocorrelation, a time series decomposition can be applied. As a third step, it can be further explored if dynamics of spatial autocorrelation coincide with relevant factors such as favourable or unfavourable socio-economic trends and policy changes. We highlight different regional labor markets in Europe, which makes for an interesting exploration due to their diversity and policy relevance.