Visualizing global international migration flows

Nikola D. Sander, Wittgenstein Centre (IIASA, VID/ÖAW, WU)
Ramon Bauer, Wittgenstein Centre (IIASA, VID/ÖAW, WU)
Guy J. Abel, Wittgenstein Centre (IIASA, VID/ÖAW, WU)

There is a long tradition in demography to display research findings using text and tables. While this approach allows data to be shown in a precise manner and provides an efficient means to look up values for a particular country or year, tables fail to highlight patterns and trends, and to tell the story behind the data. This is especially true for very complex datasets, such as migration flow tables. Despite all the progress that has been made regarding geographic information systems (GIS), migration flows are typically visualised using maps with arrows indicating the size and direction of flows. Such maps are low in complexity and visual appearance, allowing only a small number of flows to be shown. We present a new way to visualize complex migration flow data in an efficient and, indeed, beautiful manner. We draw on new estimates of 5-year bilateral flows between 196 countries from 1990 through 2010 that are estimated from the United Nations sequential stock tables by the third author. Our visualisation is open source and thus can readily be used to uncover the key patterns and trends underlying any kind of flow data, including trade and remittances flows. Specifically, we present three new methodologies for visualising flow data: First, we present circular plots created using Circos, software based on a Perl script and frequently used in genetics. Second, we show how these circular plots can be produced using R. And third, we use the Java Script library d3.js to interactively visualise global migration flows on the web, providing visually highly appealing graphics that invite the user to intuitively explore our new set of estimated global migration flows. Our novel representation has the potential to transform the way we evaluate temporal and spatial patterns in flow data.

  See extended abstract

Presented in Poster Session 2