Immigrants labour market segregation in Italy. A multilevel approach

Giuseppe Gabrielli, Università degli Studi di Napoli Federico II
Nicola Tedesco, Università degli Studi di Cagliari
Anna Paterno, Università degli Studi di Bari
Luisa Salaris, Università degli Studi di Cagliari

The global economic crisis has had important effects on international migration and employment situation of foreign workers. Nevertheless, immigrants in Italy are continuing to grow although at a slowing speed, mainly due to the recent economic crisis. The foreigners’ quota in the total population have passed from 5.7% to 7.7% in the 2007-2010 period. In the same period the foreigners in the labour force increased from 1.5 million in 2008 to 2.4 in 2010, as to say from 6.5% to 9.1% of total labour force. Recent comparative studies have shown that immigrants in Italy experience serious difficulties in entering regular and skilled jobs, whereas have fairly easy access to irregular, unskilled and semi-skilled jobs, with no chances of a future upward promotion. Moreover, inequalities between immigrants and autochthonous workers often lead to situations of labour “ethnicization” and segregation. In general, the Italian labour market is characterized by significantly different levels and patterns in occupation conditions, according to geographic areas and gender. This heterogeneity grows if we observe also the foreign presence by country of origin (Tedesco et al., 2012). The goals of this contribution are twofold: first, obtain descriptive measures of segregation in labour market according to different geographic areas and countries of origin; secondly, estimate different levels of segregation via a cross-classified multinomial multilevel model. We use the micro-data coming from the ISTAT “Continuous survey on labour forces”, observing the cohorts of 2007 and 2010. The latter analysis allows to overcome the problem of overestimation of magnitude of segregation due to descriptive indexes. In particular, estimates of variances by areas or by countries of origin through a multilevel model will provide the measures of labour market segregation stratified in different groups and/or times (Leckie et al., 2012), while estimated parameters will offer measures of associations with potential predictors.