Exploring social dynamics: predictive geodemographics

Geodemographics is the analysis of people by where they live, underpinned by the notion that people residing within proximity exhibit similar characteristics. Geodemographic classifications are typically area classifications built with composite variables, in order to segment populations into homogenous groups. However, such composite variables are often collated from static data such as population census, and other administrative and or commercial sources which are not regularly updated. Furthermore, most commercial systems are black-boxes: they do not provide information about their data, methods, updates or clusters. Consequently, this means that geodemographic classifications offer out of date and opaque representations of the populations that they seek to segment. Since such classifications have a range of public and private sector uses such as education, policing, health, targeted advertising, and location optimisation, such static data sources create many temporal limitations, and drawbacks. This research describes the social processes that suggest changes in socio-economic status and therefore geodemographic classification. Using a case study, it explores how social dynamics and geodemographic trajectories over time can be predicted from data that capture social process and enable area’s future geodemographics to be predicted. Socially dynamic processes include gentrification (area improvement, thus an improved geodemographic cluster trajectory), urban decay (area deterioration, thus a worsened geodemographic cluster trajectory), and area stability (thus a stagnated but stable geodemographic cluster trajectory). A number of future research areas are identified.