• Debraj Roy

Fractal dimension of human migration within Netherlands

Modelling the Influence of Regional Identity on Human Migration.

Geographical representation of the used NUTS 3 regions.
Geographical representation of the used NUTS 3 regions.

Human migration involves the relocation of individuals, households or moving groups between geographical locations. Aggregate spatial patterns of movement reflect complex interactions among motivations (such as distance, identity, economic opportunities, etc.) that influence migration behaviour and determine destination choice. Gravity models and radiation models are often used to study different types of migration at various spatial scales.

“While international migration has gathered more attention due to its policy implications, internal migration, with an estimated total of 740 million people (12% of the world population) in 2000, is much larger than an estimated 200 million international migrants.”

Research on internal migration is important, as it is a key aspect of the population dynamics of a country. Even though administrative structures can be hard to compare, the United Nations’ Department of Economic and Social Affairs shows that five-year migration rates between the smallest administrative units in countries worldwide can vary between 4% and 20%. In a recent paper, a team of researchers from the Computational Science Lab at the University of Amsterdam proposed that human migration models can be improved by embedding regional identities into the model. They modify the existing human migration gravity model by adding an identity parameter based on three different sets of Dutch identity regions. Through analysis of the Dutch internal migration data between 1996 and 2016, the study shows that adding the identity parameter has a significant effect on the distance distribution. The team finds that individuals are more likely to move towards municipalities located within the same identity region. They test the impact of regional identity by comparing randomly spatially clustered and optimised identity regions to show that the effects we attribute to regional identity could not be attributed due to chance. Finally, the finding shows that cultural identity should be taken into account and has broad implications on the practice of modelling human migration patterns at large. The study shows that people living in Dutch municipalities are 3.89 times as likely to move to a municipality when it is located within the same historic identity region. Including these identity regions in the migration model decreases the deviation of the model by 10.7%.

The Road Ahead

Regional identity is an important factor in the human migration decision, it would be interesting to embed these identity regions in other types of human migration models as well. When such models are used on regions in different circumstances, this could provide more information about the influence of regional identity itself. It would be interesting to see what would happen in more segregated societies, or in societies where most people regularly travel over larger distances than in the Dutch society..

Link to the full paper:

69 views0 comments