The proposed research develops and tests models of individual behavior to provide evidence on the magnitude, causes and consequences of the mass migration between Europe and the US that occurred at the turn of the twentieth century. Underlying the project is the availability of electronic administrative records for 24 million migrants who arrived in the US via Ellis Island between 1892 and 1924, that we have obtained access to. Our earlier work using this data [Bandiera et al. 2011] shows that migration at the turn of the twentieth century was effectively a two-way flow between the US and Europe, rather than a one-way movement from Europe to the US. This insight is what the proposed research agenda seeks to build on. The proposed project will develop and apply economic theory and micro-econometric methods related to core questions in the economics of migration. Our proposal will provide theory and evidence on four broad research themes: (i) the determinants of temporary versus permanent migration, and consequently how each type of migrant is differentially selected; (ii) the behavior and socio-economic outcomes of migrants who endogenously chose to remain in the US; (iii) the impact of mass migration on the labor market outcomes of Americans; (iv) whether institutional change in the US was driven by the nature of selective migration into America and where migrants chose to settle. In consequence, and to return full circle to the original insight from Bandiera et al.  that underlies this research proposal, we ask whether migrants that returned to Europe from the US drove institutional change across European countries at the turn of the twentieth century.