Abstract
Spatial decentralization is a common measure used by public planners to redistribute urban social and economic activities in order to resolve escalating problems such as crowdedness, pollution, and high cost of living in the central cities. Findings from studies assessing the changing rent gradient of cities, and probability of development around suburban centers, provide insights to the various impacts of spatial decentralization policies. The author makes use of global and local spatial auto-correlation statistics and seeks to explore the spatial clustering of property values in the context of polycentric urban development. Data were collected from the property sales transaction database in Singapore and analyzed employing geographic information systems and spatial statistics. Empirical results suggest that there was a remarkable redistribution among the planning regions as regards the number of condominium projects in the 1990s, which was in line with the polycentric urban development policy. A decline in the global Moran's I indicates a change from a strong to a relatively weak positive clustering. However, the local Moran statistics show that a spatial pattern of property-value clusters continued during the study period. Findings of this research capture a market signal indicating that the spatial distribution of property values may not readily follow planning guidelines, at least in a given short time frame.