Purpose
The purpose of this paper is to establish measurable factors that can be used as early indicators of which neighborhoods are most likely to undergo a process of gentrification in a reasonably near future.
Design/methodology/approach
Using 1990 data on key demographic variables for 128 neighborhoods in Stockholm, Sweden a model that allows both for testing of spatial clustering and for spatial spillovers between neighborhoods is estimated. It is hypothesized not only that gentrification depends on inter-neighborhood characteristics but also that gentrified neighborhoods will cluster and preferably be located in proximity to existing high income neighborhoods.
Findings
The findings confirm the stated hypotheses. Among the results, it is shown that neighborhoods that gentrified between 1990 and 2012 were more likely to have been poor in 1990 and located closer to the CBD, they were also more likely to be close to neighborhoods with lower proportions of low income residents. It is also found that gentrified neighborhoods tend to cluster over space.
Originality/value
Much of the previous literature on gentrification has concerned the actual driving forces behind gentrifying neighborhoods. This paper is more concerned with indicators that can be used to spot neighborhoods that are likely to undergo a gentrification process in the future. Such information can be valuable for real estate developers in the private sector, as it may lead to more successful investments. It may also be useful for city developers at the municipal organization.
