Finding the state space of urban regeneration: Modeling gentrification as a probabilistic process using k-means clustering and Markov models

Gentrification is a dynamic, globalized urban process whose complex definition varies with stakeholder perspectives. This complexity makes it challenging for researchers to study the impact of gentrification, and difficult for planners to anticipate the effects of gentrification with planning policy. This paper proposes to model gentrification as a Markov process, i.e. a process that assigns probabilities to potential "state" changes over time (Rabiner, 1989). Using American Community Survey (ACS) data for four boroughs of New York City between 2009 and 2013 (including demographic, economic, geographic, and physical characteristics of census block groups), we develop our model in three steps: 1) clustering census block groups into states defined by ACS socioeconomic and demographic data, 2) deriving a Markov model by tracking transitions between states over time, and 3) validating the model by testing predictions against historic data and comparing them with qualitative documentation.