Assessing the multidimensional structuring of quality of life. Evidence from a French city

This study develops a multidimensional approach to the analysis of the relationships between objective indicators and individuals' evaluation of their quality of life, in the context of a growing medium-sized French city, Bordeaux. We build a unique dataset composed of objective and subjective indicators of eight life domains for all the 303 municipalities within the urban area of Bordeaux. We then apply a variable clustering approach (ClustOfVar), an innovative multivariate exploratory statistical method, to: (1) determine distinct classes of living conditions within the regional urban system; and (2) identify which life domains matter for people and how a specific life domain is associated with other domains in contributing to different profiles of individual overall life satisfaction. The sampling rule for the survey of subjective well-being is based on the classes of municipalities' living conditions, to alleviate the problem of data mismatch when combining objective variables with individual subjective assessments. The findings show that while all the identified life satisfaction profiles can be observed in urban and peri-urban municipalities, individuals with high levels of well-being are more likely to live in periurban areas. These results are of practical significance for urban development planning and regional policy.