Neighbourhood change and spatial inequalities in Cape Town

Abstract

Capturing the dynamics of population change in urban areas necessitates access to geographically fine-grained and temporally consistent data for several time points. Such data are generally not available and they must be created using standard population data which cannot usually be compared across time periods. In this paper, the focus is on changing spatial inequalities in Cape Town, South Africa. This paper details an approach to generation of gridded population counts (250m by 250 m) for two Census years – 2001 and 2011. Census data for Small Area Layers (SALs), Spot Building Count (SBC) data, and Open Street Map (OSM) landuse data were used to construct a grid of populated cells to which population counts are then reallocated. The reallocation of population counts from SALs to grid cells was undertaken using area-to-point kriging – an approach which is informed by the spatial variation in the population groups of interest as measured using the variogram. A case study based on population grids of unemployment rates shows how the grids can be used to chart changes and also to measure spatial inequalities across the city at two time points. The advantages of grids for capturing fine-scale complexities and correctly accounting for physical separation between communities is demonstrated and the results show, while the broad patterns of inequality are consistent across time, there are pronounced increases in inequalities in some neighbourhoods. These areas – and what leads some areas to fall further behind – should be the subject of attention by policy makers.