In this paper we describe the role of ethnicity in spatial patterns of mover _flows in Scotland by focusing on the ethnicity of movers in relation to the ethnic composition of neighbourhoods. Many cities are becoming more and more ethnically diverse. Within cities however, neighbourhood population composition might not always reflect this diversity as clustering, segregation, but also mixing of ethnic groups predominantly takes place on lower spatial levels. For individual residents, the ethnic population composition of a neighbourhood might therefore become an increasingly distinctive and important feature for the comparative evaluation of residential locations. If the ethnic population composition of a neighborhood plays a role in individual location decisions, compositional preferences are most likely related to one's own ethnicity._Understanding this mechanism in the process of residential location decisions will ultimately help us to better understand ethnic neighbourhood dynamics.

For the analysis in this paper, we use a unique combination of datasets on a very fine-grained spatial level. We link the Registers of Scotland house transactions data between 1990 and 2010 with the Scottish Neighbourhood Statistics. The ethnicity of house buyers is identified by (sur)name analysis. There is a rich literature on using onomastics to identify ethnicity, and_this approach can overcome some of the drawbacks inherent in using other ways of identifying ethnicity like, for example, country of birth.

Adapting a gravity model of spatial interaction, we model spatial patterns of mover flows of different ethnic groups in Scotland and neighbourhood compositional change. Spatial interaction models offer a systematic approach to describe the role of neighbourhood population composition in determining location patterns of different ethnic groups.