Are immigrants an asset or a burden to the receiving economy? Does emigration deprive source populations of the most skilled and educated portions of their labor force, or does it provide otherwise-unavailable opportunities to the most deprived? A first step in answering these broad and fundamental questions of the economics of migration is to study the nature and determinants of the productivity of migrants in their place of origin and in their destination. This project focuses on three specific questions in this vein. First, are migrants positively or negatively self-selected from within their populations of origin on the basis of productivity? That is, are migrants generally more or less productive than stayers? Second, how is such self-selection affected by the state of the economy in the sending and receiving countries? Finally, are locality-of-origin-specific characteristics of migrants more powerful predictors of outcomes and productivity in the receiving labor market than are migrants' idiosyncratic characteristics? That is, does ability compensate for a poor background?
These questions are studied with the aid of a novel data set to be constructed by the PI. This data set will consist of the personal characteristics of Italian passengers arriving in the Port of New York between 1907 and 1925, as reported on the passenger manifests stored at Ellis Island. Individuals in these manifests will also be linked to decennial US Census manuscripts in order to recover information on their post-migration labor market outcomes, as well as the labor market outcomes of their US-born children. Three features of this data set allow the PI to improve on the shortcomings of prior studies. First, an anthropometric measure, average stature, is used to quantify migrant productivity. This measure is known to be a very good proxy on average for many variables that are strongly correlated with individual labor productivity (e.g., health, skill, and education), and is not subject to many of the problems that accompany the use of common measures of productivity, such as literacy and occupation. Second, the data enable the identification of migrants' specific place of origin within Italy, and thus their matching with and comparison to highly disaggregated stature data from Italy. It is therefore possible to compare migrants directly to their local population of origin. Such disaggregation is rarely possible with any other measure or in any other time period of study, even on the national level. Finally, these data make it possible to observe both the post-migration and pre-migration characteristics of migrants, enabling a comprehensive study of migrant assimilation.
The study of historical migration has several advantages over the study of modern migration. Migrant self-selection in modern contexts is masked by a number of legal barriers that prevent the migration of certain types of migrants. Data on modern migrants are therefore not representative of the entire population that would choose to migrate in the absence of legal barriers. Moreover, historical data are not subject to the confidentiality restrictions that constrain research with modern data, making the creation of the proposed dataset - particularly the combination of migration and census records - possible. The study of historical migration can therefore equip researchers with a better understanding of the forces driving the composition of migratory flows and the labor market outcomes of migrants in the receiving country. Results from this research can therefore provide valuable insights to policy makers that would be unattainable through the use of modern data alone.
