This was a paper that I worked on in the “Statistical Analysis of Networks” class while doing my MSc in Oxford. Final grade: 68
Summary:
In recent years, migration has emerged as one of the most controversial topics in policy circles. It remains to be a consistent theme amongst voters and politicians who are concerned with its impact on national security, labour and employment, and overall economic welfare. Critics note that migrants are mostly from developing countries, and that their attraction to the promise of higher wages and a better quality of life comes at the expense of citizens whose jobs and economic resources are “being taken away” (Hoban, 2017). Another observation is that immigrants hail from conflict-ridden, war-torn, and fragile nations which may pose security risks to the destination country (Koser, 2001). My goal with this paper is to use statistical analysis to investigate whether these two themes emerge in migration networks. Is migration to the developed world indeed driven by citizens from developing countries and conflict states?
By analyzing network motifs – or overrepresented small graphs in the immigration network, I observe patterns in the topology and structure of the network, which could then provide information on the behaviour and flow of migrants globally. I particularly focus on one class of motifs: the triad, or ordered triples that are generally mapped to 16 types.
Comparing the immigration network to a directed and unweighted configuration model, I find that three motifs emerge (210, 120D, and 300). An analysis of the countries in these motifs suggest that immigration to high-income economies remains elusive to lower-income countries.If anything, high-income economies enjoy the mobility to migrate freely to other countries, while lower and middle-income economies do not appear to have the same benefit. There is also very little evidence that points to substantial migration from fragile states, though it seems that the weighted network may provide clues on an emerging trend. Thus, I suggest that in a future analysis, one may explore the possibility of weighted motifs in order to better understand the dynamics of migration flows.
Code: https://github.com/carlaint/OII_Statistics_of_Networks
Full version of the paper: message me on LinkedIn!
