Measuring and mapping long-term changes in migration flows using population-scale family tree data

Abstract

Studying migration over a long period is challenging due to lack of data, uneven data quality, and the methodological challenges that arise when analyzing migration over large geographic areas and long time spans with constantly changing political boundaries. Crowd-sourced family tree data are an untapped source of volunteered geographic information generated by millions of users. These trees contain information on individuals such as birth and death places and years, and kinship ties, and have the potential to support analysis of population dynamics and migration over many generations and far into the past. In this article, we introduce a methodology to measure and map long-term changes in migration flows using a population-scale family-tree data set. Our methodology includes many steps such as extracting migration events, temporal periodization, gravity normalization, and producing time-series flow maps. We study internal migration in the continental United States between 1789 and 1924 using birthplaces and birthyears of children from a cleaned, geocoded, and connected set of family trees from Rootsweb.com. To the best of our knowledge, the results are the first migration flow maps that show how the internal migration flows within the U.S. changed over such a long period of time (i.e. 135 years).