Understanding the spatial patterns of social inequalities has been a longstanding concern in urban studies. Geodemographic classifications, which group neighbourhoods based on multiple social and physical dimensions, offer a useful tool for this purpose. However, most classifications rely on fixed single-scale administrative boundaries, while studies that adopt multiscale approaches often focus on a single dimension and cover only limited time periods. This limits our understanding of how urban social inequalities evolve over time and across spatial scales. In this study, we extend the geodemographic approach to incorporate multiple dimensions, time periods, and geographical scales, enabling a more comprehensive analysis of the spatio-temporal configuration of urban change. We develop multidimensional, multiscale, and longitudinal spatial profiles of residential contexts in the Metropolitan Agglomeration of Amsterdam (MAA) using bespoke neighbourhoods constructed from detailed population register data (1999–2022). Our results show that the interaction of socioeconomic status, migration background, life-course stages, and housing tenure provides a richer understanding of urban stratification than traditional models based solely on income or ethnicity. The longitudinal perspective reveals distinct timing differences in urban reconfigurations, such as gentrification and displacement, which emerge locally and consolidate more broadly over time. The multiscale approach highlights how patterns of urban change are scale-dependent, with large-scale dynamics, such as poverty suburbanisation and inner-city gentrification, coexisting with the formation of smaller enclaves in areas undergoing or at risk of change. These findings highlight the need for integrated multidimensional, temporal, and multiscale frameworks to better capture the evolving nature of sociospatial inequalities in cities.

Traditionally, studies of spatial inequalities only consider one single dimension, such as income, and one spatial scale - usually a neighbourhood determined by administrative boundaries. Although the existing literature increasingly recognises the multifaceted nature of inequalities in cities, this paper introduces a novel approach by integrating the multidimensional and multiscale perspectives to understand the evolution of social and spatial inequalities over time. Drawing on clustering techniques based on factor analysis and using individual-level geocoded register data from the metropolitan agglomeration of Amsterdam, our methodology classifies neighbourhoods by grouping detailed residential locations with similar socioeconomic, demographic and housing characteristics across multiple geographical scales. Through sequence analysis, we identify trajectories of neighbourhood change from 1999 to 2022, revealing patterns in the timing, duration, and sequencing of shifts across various dimensions. Our results bridge gaps in the multidimensional and multiscale neighbourhood classification literatures, providing a better understanding of how social inequalities interact and overlap in space. By examining the path dependence between different dimensions of spatial and social inequalities, this study provides insights into the processes that produce and reproduce social stratification in cities that may act at different geographical scales for different groups of people. Moreover, the rich and granular data paint a detailed picture of how residential contexts are segregated and how the trajectories of neighbourhood change are distributed spatially. This research offers an innovative framework for visualise and study the dynamic evolution of urban structures over time.

It has been widely documented that both housing conditions and household overcrowding over time negatively affect physical and mental health. However, scant evidence documents this dynamic relationship in the low- and middle-income countries of Latin America, where housing issues remain a relevant policy consideration. Employing a nationally representative panel dataset of 10,024 Chilean households, we examine whether variation in household overcrowding levels between 2006 and 2009 is associated with the prevalence of depressive symptoms in 2009, measured by the Centre for Epidemiological Studies Depression Scale (CES-D). The survey allows us to follow households over time to identify those who have experienced variation in overcrowding due to changes in household size or due to changes in housing conditions, which do not occur too frequently over time. We find that an increase in household overcrowding levels (due to a reduction in the number of available bedrooms) is associated with an increase in depressive symptoms, while a constant or decreasing trajectory of household overcrowding over time is not associated with changes in depressive symptoms. These results suggest an asymmetric relationship between household density and mental health over a three-year window and highlight the importance of preventive housing policies to address overcrowding alongside policies to reduce existing overcrowding.