Multidimensional Analysis of Deprivation and Fragility Patterns of Migrants in Lombardy, Using Partially Ordered Sets and Self-Organizing Maps

Abstract

In this paper, we present a multidimensional fuzzy analysis of the levels and the patterns of poverty and social fragility of migrants' families, in the Italian region of Lombardy, in year 2014. Migrants' poverty emerges as a complex trait, better described as a stratification of nuanced patterns than in black and white terms; Lombard migrants are in fact affected, to different extents, by "a diffused sharing of deprivation facets" and cannot be trivially split into deprived and non-deprived. The paper employs innovative data analysis tools from the Theory of Partially Ordered Sets; compared to mainstream monetary approaches, this leads to more realistic estimates of poverty diffusion and eliminates some well-known biases of standard evaluation procedures, providing strong support to the use of partial order concepts and tools in social evaluation studies.

Targeting Policies for Multidimensional Poverty and Social Fragility Relief Among Migrants in Italy, Using F-FOD Analysis

“…In a recent paper by Arcagni et al (2019), multidimensional poverty and social fragility of migrants' families in Lombardy (Italy) have been thoroughly explored, revealing a quite complex pattern in the levels and shapes of their social conditions. By using concepts from partial order theory, the study provides realistic estimates of poverty diffusion and sheds light on the social differences between and within migrants' groups, indirectly implying that no poverty relief policy can be effective, without accounting for the nuances of such inhomogeneous scenario.…”

“…We differ to a later section the description of such a statistical tool; here it suffices to say that F-FOD allows for multidimensional statistical distributions defined on partially ordered sets to be compared, in terms of relative dominance, without preliminarily "collapsing" them into some kind of synthetic indicators. In its essence, F-FOD performs a "synthesis of multidimensional comparisons between distributions", rather than performing a "comparison of synthesized multidimensional distributions", as one would do by using more classical aggregative procedures, like the much-revered Counting Approach of Alkire and Foster (2011a, b), or even the non-aggregative posetic approach to poverty evaluation, employed in Arcagni et al (2019). This way, the dimensionality reduction step is "postponed" to the end the statistical process, making F-FOD inherently more "information and complexity preserving" than procedures passing through the evaluation of poverty or fragility at individual level.…”

“…This way, the dimensionality reduction step is "postponed" to the end the statistical process, making F-FOD inherently more "information and complexity preserving" than procedures passing through the evaluation of poverty or fragility at individual level. In this respect, the F-FOD procedure extends the classical univariate stochastic first-order dominance criterion, to the multidimensional ordinal setting, typical of social evaluation studies, and meets the need for more suitable poverty measurement tools; in particular, it helps overcoming the classical approaches based on monetary poverty lines, in particular when focusing on foreigners (Arcagni et al 2019;di Belgiojoso et al 2009;Busetta 2016;Lemmi et al 2013).…”

“…This motivates the focus of the present paper, which is specifically devoted to the ranking of migrants' families, for policy targeting. More concretely, we cluster migrant families by country of origin and years since migration, two features that have been observed to be linked to migrants' social conditions (Arcagni et al 2019), and compare the corresponding frequency distributions on a multidimensional set of poverty and fragility indicators, by using the aforementioned F-FOD procedure. From the resulting set of pairwise comparisons, a final ranking is eventually obtained, identifying the most critical subpopulations, to be considered by policy-makers.…”

On the Interaction Between Focus and Distributional Properties in Multidimensional Poverty Measurement

“…See, for instance, Chakravarty and D'Ambrosio (2006), Alkire and Foster (2011), Bossert et al (2013) and Aaberge and Peluso (2012). In turn, Fattore (2016) proposes an operative procedure for assessing multidimensional poverty in the presence of ordinal attributes, which is based on partially ordered sets and avoids the need of aggregation; Arcagni et al (2019) offer an empirical application of such approach. vation focus property include the ones proposed by Bresson (2009), Decancq et al (2014), and one of the three measures proposed by Maasoumi and Lugo (2008).…”