Bridging across nations. The social capital of diversity, brokerage and closure in transnational migrant networks: a study on assimilation patterns in Milan and Barcelona

This dissertation deals with the network effects on immigrant assimilation: how the composition and structure of migrants' personal networks affect patterns of assimilation in host societies. A personal network is the set of all current and active social contacts that a focal individual (Ego) has, including family, friends and acquaintances. It is operationalized here as the list of 45 alters who were mentioned by Ego (the migrant) as people with whom he had some kind of contact in the last two years, and whom he could currently still contact. In particular, personal contacts can be of any nationality, and live in any country. In network terminology, ego-network data are used here. They come from two separate surveys on migrant personal networks and assimilation patterns, carried out in Barcelona, Spain, and Milan, Italy. The Spanish survey collected data on 289 international migrants from Argentina, Dominican Republic, Morocco, Senegal, and Gambia, in the years 2005-2006. In the Italian survey, 107 immigrants from Sri Lanka were interviewed in 2011-2012. The author managed the latter survey, which was prepared by extensive ethnographic work in Milan and Sri Lanka. Migrant personal networks are analyzed in both their composition, that is, the distribution of actor attributes in the network, and structure, that is, the pattern of relations in the network. Most analyses are based on two attributes of network actors: nationality and country of residence. A step further is proposed, beyond the traditional dichotomy between network composition and structure, with a study of the pattern of relations between different actor attributes. This combines the usually separate dimensions of composition and structure to examine the structural cohesion versus segregation of actors from different nationalities, or in different countries of residence. A specific measure is proposed to index cohesion versus segregation of different attributes in the network, which is called the index of subgroup segregation: “geographical” segregation, if it concerns actor countries of residence; “national” segregation if it regards actor nationalities. Theoretically, the index of subgroup segregation can be related to the popular sociological concepts of structural brokerage and network closure. However, it describes a particular kind of brokerage and closure: brokerage and closure between actors from different nationalities and in different countries of residence, which are called “brokerage between differences” and “diversity within closure” in the text. Empirically, the index of subgroup segregation is shown to capture an existing variation in the data, and to extract new, non-redundant information which cannot be predicted by traditional compositional and structural measures alone. Personal network metrics are used to operationalize immigrant transnationalism in particular, and investigate its relationship with assimilation. Personal networks capture different degrees of transnationalism, through compositional measures; and different types of transnationalism, through the index of subgroup segregation. Specifically, the notions of “brokerage between differences” and “diversity within closure” correspond to two substantially different types of transnationalism: transnational brokerage and transnational cohesion. Two dimensions of assimilation, or adaptation, in host societies are taken into consideration: the cultural and the economic. The former is measured by an acculturation rating scale, and studied with data on Moroccan and Senegambian immigrants in Barcelona. The latter is indexed by measures on employment status and income, and studied with data on Sri Lankan immigrants in Milan. The results strongly support the hypothesis that the meso-level of personal networks has a specific effect on assimilation, independently and separately from the micro-level of immigrants' individual characteristics. The composition of personal networks is relevant, with national and geographical diversity positively affecting assimilation outcomes. The structure of personal networks is also significantly associated with assimilation, but this effect only emerges when structure is combined with composition through the index of subgroup segregation: brokerage between differences has a positive impact on cultural assimilation; diversity within closure facilitates economic assimilation. Network transnationalism is shown to be a pervasive and still variable phenomenon among international immigrants. Results suggest that the degree of transnationalism is significantly associated to immigrant assimilation, yet in a non-linear, non-monotone relationship (first positive, then negative), similar to an inverted U curve. The type of transnationalism has a relevant impact on assimilation too: transnational brokerage fosters cultural assimilation, whereas transnational cohesion has positive effects on economic assimilation.