Abstract:
This research describes a simple, abstract model to simulate gentrification from both supply and demand side perspectives. Three theories—rent gap theory, filtering theory and household life cycle theory—are employed to construct a combined cellular automaton and agent-based model. The model exhibits a distinctive relationship between the spatial dynamics of gentrification patterns and different rent gap thresholds: at low rent gap thresholds, renovation events occur at all locations leading to a mixed rental map distribution. As the rent gap threshold increases, gentrification becomes more spatially concentrated, leading to spatially segregated rental patterns. Meanwhile, depending on both household entry and exit probabilities and the rent gap threshold, household income distributions exhibit diverse patterns: scattered and mixed at the low rent gap threshold through to segregated at a high rent gap threshold. The abstract gentrification model is evaluated against the spatial gentrification pattern at a Point Chevalier (inner Auckland) study area. Local spatial association analysis suggests that the dynamics of local gentrification in Point Chevalier mirror those of the model at a low threshold rent gap, substantiating the model’s credibility, at least in this study area. This study indicates that a different threshold rent gap is able to lead to either a mixed or segregated rental pattern. This provides a possible explanation of the dissonance in the empirical spatial gentrification pattern − some scholars observe intensified segregation after gentrification, while others do not.