Spatial Structure and Subjective Well-Being Across the South of Europe

This research examines the relationship between urbanisation and subjective well-being in the South of Europe. Its main objective is to further the understanding on this relationship at the regional level – which has been referred to as the ideal spatial level at which to study well-being (Aslam and Corrado, 2012) – by borrowing measures of spatial structure commonly used in economic geography literature, such as polycentricity, dispersion and urban size. The study thus seeks to answer one main research question: “how are different regional spatial structures associated with subjective well-being, in the South of Europe?”. This is tested in the Southern countries of Greece, Italy, Portugal, and Spain, which put forward a particular context of consistently low-levels of individual well-being, hard hit economies due to the Great Recession, and traditionally compact and monocentric towns and cities. In order to answer this question, two datasets were combined: the European Social Survey, which gathers data on participants’ socio-demographics, and the Eurostat City Statistics database, which collects city population data at the NUTS 3 level. The former provides data on respondents’ well-being, whilst the latter is used to create spatial structure indexes for polycentricity, dispersion and urban size, for a total of 48 regions. These are then tested through a multilevel modelling strategy, which assumes that regional characteristics are correlated with individual well-being levels. The study finds that life satisfaction is positively associated with regional urban size, and negatively associated with urban residence. This means that, on average, residents in these Southern European countries tend to be happier in rural residences, within more urbanised regions. However, although these results present significant associations, the effects on life satisfaction are relatively small when compared to individual characteristics such as unemployment or bad health. No significant associations were found between the polycentricity and dispersion measures, which may be attributed to the high levels of heterogeneity between regions, or the lack of a mediating functional relationship (such as commuting patterns) between individuals’ daily activities and regional structure.