Marina Toger (p), Umut Türk, John Östh, Manfred M Fischer

Thanasis Ziogas

This study investigates integrating GSM-based mobility flow data from extended Call Detail Records (CDR) into spatial econometric modeling to analyze movement patterns in the Greater Stockholm area. Using data collected over five years (2019–2023, focusing on Thursdays in March), we construct origin-destination (OD) matrices at multiple spatial scales to assess mobility trends and their socioeconomic determinants.
Initial 1,000-meter grid-level modeling resulted in sparse OD matrices and socioeconomic covariates for five years, limiting analytical depth. To address this, dynamic covariates from the demographic statistical areas (DeSO) level were aggregated to the regional statistical areas (RegSO) level, ensuring alignment between mobility flows and key socioeconomic variables, such as employment rates. This aggregation also enabled the construction of spatial weight matrices based on geographic and socioeconomic connectivity, essential for spatial econometric analysis.
To evaluate the reliability of mobile phone-generated mobility data, we benchmarked it against official municipal-level statistics, demonstrating its potential for disaggregating highly aggregated datasets into finer spatial resolutions. Beyond methodological improvements, the study explores (i) the spatial structure and temporal stability of mobility flows and (ii) the relationship between mobility dynamics and socioeconomic factors. An initial exploratory analysis using Geographic Information Systems (GIS) provides descriptive insights into mobility structures, supporting the advancement of spatial econometric modeling techniques.
The findings highlight the value of integrating CDR-derived mobility data into spatial analysis, offering new opportunities for understanding urban movement patterns and their socioeconomic drivers.

Thanasis Ziogas (p), Dimitris Ballas, Manos Matsaganis, Anastasia Panori, Jianhua Zhang

Umut Türk

Migration and all other types of human mobility decisions depend on both individual and contextual characteristics. Recently, the role of digital and green factors in shaping these decisions has been stressed through discussions focused on the geographical impact of the Twin Transition. However, data limitations at the NUTS2 level restrain the ability for performing empirical research in this field. This paper draws on findings from the EU Horizon MOBI-TWIN project which aims at addressing this gap by performing an in-depth analysis of the factors affecting individual decisions in relation to location choices, integrating traditional, digital, and green factors. Using unique individual-level data from the MOBI-TWIN survey, we estimate the probability of relocation for short, medium, and long-term periods using a multinomial logistic regression model while accounting for both individual and contextual characteristics. We then employ a spatial microsimulation model, in order to create synthetic population weights which we subsequently use to update the relocation probabilities for different socio-economic subgroups that were identified in the previous step using regression analysis. This approach enables us to gain insights in relation to the socio-economic profile of movers as well as the composition of the origin areas.

Umut Türk (p), John Östh, Karima Kourtit, Peter Nijkamp

Manos Matsaganis

This paper presents an operational statistical and modeling framework for analyzing and mapping spatial accessibility patterns among functionally overlapping X-minute cities within the national settlement system of the Netherlands. It seeks to extend the notion of local proximity – a central theme in the 15-minute city concept – toward interconnected urban areas by adapting Walter Christaller’s central place theory for the X-minute city paradigm. This neo-Christallerian approach is applied and tested using an extensive database, employing quantitative techniques such as quantile regression and hierarchical clustering to create geographic accessibility maps based on specified X-minute thresholds. The study shows significant disparities in accessibility across urban and rural regions, with urban centers like Amsterdam, Rotterdam, and Utrecht demonstrating higher accessibility scores compared to their rural counterparts. The analysis also highlights inequalities faced by vulnerable populations, particularly the elderly, in accessing essential services. GIS-based visualizations demonstrate the quality of life and spatial sustainability profiles of the Dutch urban system, emphasizing the role of active or micro-mobility in enhancing access to key amenities. Finally, the findings inform targeted policy interventions aimed at improving urban connectivity, promoting mixed-use development, and fostering active mobility to create more inclusive and resilient urban environments in the context of the X-minute city framework.

Francesco Figari, Marcello Matranga, Francesco Cellino, Manos Matsaganis (p)

Marina Toger

The paper showcases a new spatial microsimulation model, developed as an extension of EUROMOD (the tax-benefit model for the European Union), which allows us to evaluate the fiscal and distributional impact of public policies at local level. Adding spatial detail to microsimulation requires microdata reflecting the characteristics of individuals and households in particular regions, cities, or neighbourhoods. Given the lack of geographically disaggregated microdata at sub-national level, spatial microsimulation combines small area census data, tax return data in aggregate form, and survey microdata at the national or at best regional level, in order to simulate a synthetic population whose characteristics mirror as closely as possible those of the local communities under examination. We use EU-SILC 2019 data to create representative samples at local level by modifying the sampling weights based on information derived from population census and income tax returns. We apply the model to assess the anti-poverty effects of child benefits in Greece (Επίδομα παιδιού) and Italy (Assegno Unico Universale) at local level. We assess such effects by reference to two poverty benchmarks: a relative and an extreme poverty threshold set at 60% and 40% of median equivalent income respectively. We also present results in terms of absolute poverty in Italy, by reference to a threshold determined by the country’s national statistical agency (Istat), set equal to the minimum level of consumption expenditure deemed necessary “to avoid severe forms of social exclusion”. Our results reveal considerable heterogeneity at local level, and inform the debate on the optimal design of public policies.