How perceived well-being determinants differ for immigrants and natives in Italy

Happiness and well-being are thus not reducible only to physical hedonism, for they can be derived from the attainment of goals or valued outcomes in varied contexts.

Among these, the emotional component of well-being is characterized by markedly hedonistic sub-components, referring to sensations of well-being or physical and/or psychological disease.

Applying BO decomposition in socio-economic studies, the coefficients effect is often interpreted as an effect of “discrimination” due to the subject belonging to one of the two groups. It is the case, for example, if the two groups are represented by males and females, or by blacks and whites (e.g., Oaxaca 1973).

This phenomenon should be further investigated in a specific study. Note that the introduction of an additional age-squared coefficient - with the aim of reproducing the well-known “U-shaped” relationship between happiness and age in the model - did not significantly affect estimation results or the conclusions of our analysis. The introduction of age groups (young people, adults, seniors) did not produce significant differences. On the other hand, also due to the cross-sectional structure of the dataset, the age effect is difficult to separate from the period and cohort effects and, to a certain extent, from the income effect. Note, on this point, a positive and significant impact of age on income obtained from the estimation of the reference income function (Table 3). Interesting discussions on the U-shaped’s relationship between age and happiness can be found in Frijters and Beatton (2012), Graham and Ruiz Pozuelo (2017) and Beja (2018).

Performing the BO decomposition model, we used macro-areas as a covariate, not regions. The subdivision by region allows us to perform a post-estimation differential analysis on SUR estimates, by calculating the regional averages of SUR predicted values.

Another possible approach is to build a synthetic index including different components of well-being. A recent interesting example is in Sironi (2019), where ten items are used to measure optimal well-being and grouped to build three categorical indicators: positive emotion, positive characteristics, and positive functioning. A synthetic well-being index is built combining these three indicators.