This article examines the role of migrant social networks in international migration and extends prior research by testing the strength of tie theory, decomposing networks by sources and resources, and disentangling network effects from complementary explanations. Nearly all previous empirical research has ignored friendship ties and has largely neglected extended-family ties. Using longitudinal data from the Migration between Africa and Europe project collected in Africa (Senegal) and Europe (France, Italy, and Spain), this article tests the robustness of network theory-and in particular, the role of weak ties-on first-time migration between Senegal and Europe. Discrete-time hazard model results confirm that weak ties are important and that network influences appear to be gendered, but they do not uphold the contention in previous literature that strong ties are more important than weak ties for male and female migration. Indeed, weak ties play an especially important role in male migration. In terms of network resources, having more resources as a result of strong ties appears to dampen overall migration, while having more resources as a result of weaker ties appears to stimulate male migration. Finally, the diversity of resources has varied effects for male and female migration.

This paper examines the gendered roles of sibling position and network-derived social capital in Mexican and Senegalese international migration. We investigate how men's and women's migration decisions are associated with their position within the nuclear family before and after accounting for nuclear family migrant networks. Crucially, we also estimate how sibling network "effects" are gendered. We analyse 2 comparable household surveys in very distinct settings where family obligations may vary: the Mexican Migration Project (1998)(1999)(2000)(2001)(2002)(2003)(2004)(2005)(2006)(2007)(2008)(2009)(2010)(2011)(2012) and the Migration between Africa and Europe-Senegal project (2008). Using discrete-time event history analysis, we find that-without controls for nuclear family migrant networks-younger siblings in Mexico appear more likely to migrate than their older counterparts, whereas the opposite is true in Senegal. After we control for family migrant networks, however, older siblings in both countries and of both sexes are more likely to migrate. Despite these commonalities, migrant sibling networks appear to be gendered differently in the 2 countries. Although Mexican men are most influenced by migrant brother networks and Mexican women are most influenced by sister networks, evidence for Senegal is mixed and destination-dependent. Brother networks motivate Senegalese men's and women's migration to Europe, whereas nearly all sibling networks, regardless of gender, influence Senegalese migration within Africa.

BACKGROUND Longitudinal micro-level data about international migration behavior are notoriously difficult to collect, but data collection efforts have become more frequent in recent years. Comparative research of the patterns and processes of international migration, however, remains quite rare, especially that which compares across regions. OBJECTIVE We highlight the promises and difficulties of comparative international migration research, by offering a detailed comparison of two prominent data collection efforts. METHODS We systematically review existing sources of longitudinal and quasi-longitudinal individual-level and household-level data of international migration. We then compare two widely-used data sources: the Mexican Migration Project (MMP) and the Migration between Africa and Europe project (MAFE). RESULTS Data collection efforts are increasingly diverse, yet public accessibility of data remains limited. Also, comparability of data collected across settings can be complicated. In our MMP-MAFE analysis, we show some ways in which comparability can be achieved. CONCLUSIONS A primary roadblock to international comparative research is that, with some exceptions, the public accessibility of data remains low. Even when data is public and surveys are modeled after one another, comparability is not easy due to necessary trade-offs in adapting surveys to local settings and to developments in the field. CONTRIBUTION We demonstrate that, despite great strides in collecting quasi-longitudinal data of international migration, data accessibility still hinders the study of migration. With regards to comparability, our article provides important lessons for future data collection and analysis efforts that could improve comparability and thus advance understanding of the complex dynamics of international migration.