Building Well-Being Composite Indicator for Micro-Territorial Areas Through PLS-SEM and K-Means Approach

• Published in: Social indicators research Vol. 153; no. 2; pp. 407 - 429
• Main Authors: Tomaselli, Venera, Fordellone, Mario, Vichi, Maurizio
• Format: Journal Article
• Language: English
• Published: Dordrecht Springer Netherlands 01.01.2021; Springer Nature B.V
• Subjects: Classification; Clustering; Environmental quality; Happiness; Human Geography; Least Squares Statistics; Microeconomics; North and South; Original Research; Public Health; Public policy; Quality of life; Quality of Life Research; Relative deprivation; Social Sciences; Sociology; Statistical methods; Structural equation modeling; Structural Equation Models; Sustainability; Well being; Italy; Sustainable well-being; Composite indicators; Partial least squares-structural equation models; K-means clustering
• ISSN: 0303-8300; 1573-0921; 1573-0921
• DOI: 10.1007/s11205-020-02454-0

In the analysis of the difference in the distribution and profiles of the equitable and sustainable well-being, the territorial dimension is a fundamental reading-key for local policies since it allows the areas of advantage or relative deprivation to emerge more accurately. Specifically, in Italy the provincial level coincides with the administrative area of metropolitan cities, which are the subject of growing attention from European and national policies. The BES 2018 report by Italian National Institute of Statistics (ISTAT) has confirmed that from 2015 an improvement in many areas of well-being has been marked, even if territorial differences remain stable both in levels and dynamics. These differences appear in some cases as real structural differences between the North and South of Italy. Then, the measures of equitable and sustainable well-being in the territories allow, in various degrees, to deepen and specify this situation employing synthetic measures of well-being. In this work, we propose a statistical methodology focused on the simultaneous partial least squares structural equation modeling and simultaneous K-means clustering to obtain a composite indicator of Italian well-being and at the same time a classification of Italian territorial micro-areas by means of the just updated provincial data about BES 2018. In this way, the territorial differences of well-being can be more reliably and more exactly defined on the basis of the relationships among all elementary indicators and domains proposed in the analysis of well-being by ISTAT.