The role of individual characteristics and municipalities in social
inequalities in perceived health (Italy, 2010–2012): a multilevel study
Received: 11 April 2018 / Accepted: 18 May 2018
Springer-Verlag GmbH Germany, part of Springer Nature 2018
Backgrounds The empirical evidence shows discordant results regarding the role of local contexts on individual health. This
article considers the role of the municipal socio-economic contexts on self-rated health in Italy, taking into account some
Methods Multilevel model software (MlwiN) is used to fit multilevel linear regression models of perceived health. Individual
data are from the Italian surveys on BAspects of Daily Life^ 2010, 2011 and 2012, collected by the Italian National Institute of
Statistics (Istat). In addition, municipality-level social, demographic and economic characteristics are from the 2011 Census and
the database BAtlas of Italian Municipalities^ (Istat).
Results The main findings of this study confirm that, controlling for age and gender at the individual level, poor health is
influenced by socio-economic positions: lower education, not working or looking for employment and disadvantaged family
social class predict higher perceived health. The individual level explains the 70.1% heterogeneity in self-assessed health, the
family level 25.6% and the municipality level only 4.3%. The additional influence of the socio-economic context is, conversely,
of little substantive importance.
Conclusions Finally, by showing that variability in health relates mainly to individual characteristics, this study suggests that
intervention to mitigate social inequalities in health should focus on structural factors, such as education and the labour market.
Inequalities in health
The health of the population is an issue of critical importance in
human societies that affects the cultural, political and economic
spheres. Good health allows individuals to reproduce not only
biologically but also in a cultural sense through the development
of stable and well-fitting social identities (Haslam et al. 2009).
Improving the health of all human beings is one of the most
important goals declared by international organizations
(Marmot 2015). The economic burden of protecting and promot-
ing health in modern societies is an important aspect of public
and private expenditure. For example, in Western countries, ap-
proximately 10 % of the GDP is devoted to health, and this
expenditure is steadily increasing (World Bank 2017).
When health is associated with social conditions that are
attributable, for any reason, to forms of inequality, then one
can speak of social inequalities in health. From this perspec-
tive, variability in health is seen as reflecting structural disad-
vantages with regard to social organization, such as the strat-
ification of the educational system or the segmentation of the
labour market (Sarti and Zella 2016;Phelanetal.2010;
Mackenbach et al. 2008;Marmot2015;Marmot2005).
The study of social inequalities in health recognizes that the
relationship between social conditions and health may be af-
fected by several confounding variables that may eliminate or
reduce the role of social conditions. These variables can be
classified into two main interacting groups: biological and con-
textual. Moreover, we can add individualized or idiosyncratic
behaviours (self-determined on the basis of free choices) and
stochastic effects. However, these factors remain part of the
unexplained variance within the statistical models employed.
* Simone Sarti
Department of Social and Political Sciences, University of Milan, Via
Conservatorio, 7, 20122 Milan, Italy
ISTAT Sede territoriale per la Lombardia, ISTAT–National Institute
of Statistics, 20124 Rome, Italy
Journal of Public Health: From Theory to Practice