Multidimensional well-being

Working Paper 2016-399


In the multidimensional well-being literature, it has been long advocated that it is important to consider how the different well-being domains interact. Nevertheless, none of the existing approaches is useful to tackle this issue. In this paper, we show that the statistical technique of Bayesian Networks is an intuitive and powerful instrument that allows to graphically model the dependence structure among the different dimension of well-being. Moreover, Bayesian Networks can be used to understand the effectiveness of given interventions addressed to one or more dimensions, as well as to design more effective policies to reach the desired outcome. The new approach is illustrated with an empirical application based on data for a selection of Western and Eastern European countries.

Authors: Lidia Ceriani, Chiara Gigliarano.

Keywords: Multivariate analysis, directed acyclic graphs, probabilistic inference, well-being
JEL: I31, O47, O57.