This application studies the link between market imperfections and social networks. Since newcomers into an economy face particularly severe information problems, an association with migration is also obtained. The empirical analysis tests the hypothesis that network effects are present and also evaluates the performance of the networks in terms of their ability to improve market outcomes. Two applications are considered: labor-market networks set up by communities migrating from Mexico to the U.S. and a capital-market network set up by a local community in the South Indian town of Tirupur which provides capital to migrants with limited resources. Testing for network effects is a challenging empirical exercise since omitted determinants of the individual's decision or outcome are very likely to be correlated across the members of the network, giving rise to a spurious correlation. A number of alternative estimation strategies are proposed to account for such potential bias. To study the demand for health-related and other public services it is necessary to build on this framework and endogenize the assimilation process. Extensions to the analysis take a first step in this direction, linking the structure of the Mexican networks, which is determined by the historical pattern of migration, to English-language proficiency among their members.
|Eeckhout, Jan; Munshi, Kaivan (2010) MATCHING IN INFORMAL FINANCIAL INSTITUTIONS. J Eur Econ Assoc 8:947-988|