Arbeitsbereich
Migration und Mobilität
Auf einen Blick
Projekte
Publikationen
Team
Projekt
Zeiteffekte in der Messung von Migrationsbewegungen: Einsichten aus der Nutzung georeferenzierter digitaler Spuren
Emilio Zagheni; in Zusammenarbeit mit Guy Abel (Shanghai University, China), Johnathan Wayne Hill, Lee Fiorio (beide: University of Washington, Seattle, Vereinigte Staaten), Gabriel Pestre (Data-Pop Alliance, New York, Vereinigte Staaten), Emmanuel Letouzé (Massachusetts Institute of Technology, Cambridge, Vereinigte Staaten), Jixuan Cai (The Chinese University of Hong Kong, China)
Ausführliche Beschreibung
Migration data is scarce compared to data on other fundamental demographic processes, such as fertility and mortality. What is more, the migration data that do exist often measure migration under different temporal definitions, making data difficult to compare even in the same research context. These differences in temporal time scales often reflect different needs of the various institutions that collect migration data, and they reflect the different means by which such data are collected. A receiving country might classify a newcomer as an immigrant after twelve months of residency, whereas the sending country might classify the same person as an emigrant after just three months away. Similarly, an annual survey might use a one-year retrospective interval to estimate the number of migrants, whereas a decennial census might use a five-year retrospective interval.
The issues in reconciling migration data stem from the rich complexity of population movement. Unlike fertility, migration is not necessarily characterized by a singular event. Migrants might begin as temporary visitors, or they might split their time between multiple locations before ultimately settling in a new location. Furthermore, unlike mortality, migration offers no guarantee of permanence. Even a person who migrated long ago might someday return or move some place new entirely. The fuzzy boundary between short-term mobility and long-term migration has long been recognized in the migration research literature, as have been the measurement difficulties created by return and onward flows. However, the lack of high-quality data has meant that empirical research into these issues has been limited.
To address these long-standing issues in migration scholarship, this project leverages the unique spatial and temporal granularity of geo-referenced digital trace data to measure population flows at multiple timescales. In our digital era, geo-referenced digital trace data has become increasingly ubiquitous — cultivated and captured as metadata when individuals make calls and send texts or when they interact with web and smartphone applications. The structure of these data is generic, each record consisting of a tuple (containing the user ID, a time stamp, and the location of the user), and their size allows for the estimation of many different measures of migration under many different temporal specifications. In this project, by systematically varying the temporal specification, we analyze changes in migration estimates along a quasi-continuous timescale, analogous to a survival function, to evaluate data quality, assess migration patterns, and map the relationship between different kinds of estimates. Earlier research has used these new kinds of digital data to reproduce standard measures of migration. But this project goes one step further: It charts a new research agenda for improving migration measurement and advancing migration theory.
Daten und Erhebungen, Migration
Europa, Senegal, Vereinigte Staaten
Publikationen
Fiorio, L.; Zagheni, E.; Abel, G. L.; Hill, J.; Pestre, G.; Letouzé, E.; Cai, J.:
Demography 58:1, 51–74. (2021)
Fiorio, L.; Zagheni, E.; Abel, G. L.; Hill, J.; Pestre, G.; Letouzé, E.; Cai, J.:
MPIDR Working Paper WP-2020-024. (2020)