Trends of segregation by geographical scale since the early 1990’s

Aurélien Dasre, Institut National d'Études Démographiques (INED)

Quantitative measurement of segregation phenomena often imply selecting a specific geographical scale. This selection can be chosen in function of data disponibility or due to a scientific choice made by the researcher. In this presentation, we would like to combine multiple geographical scales of analyses to provide a synthetic history of the segregation trends in the biggest French urban areas. To do so, we use a specific methodology based on the decomposition of the classical Duncan segregation index, which allows identifying the implication of three different geographical scales on the evolution of job segregation in France.