Using small statistical units to abstract commuting networks in Reykjavík, Iceland

Gísli Pálsson, Ómar Harðarson, Violeta Calian – Statistics Iceland
The data synthesis underlying a national census offers many insights into the spatial character of society. For instance, joining demographic and labour force data sources allows for highly detailed mapping of likely commuter routes for employees travelling to and from work. Such commuter networks show the daily rhythms of a society in motion and can help inform city planners of the need of its citizens. This includes anticipating the needs for transport infrastructural expansion and maintenance, designing effective public transport routes and informing city planning policies by looking at where individuals are likely to spend their working hours.
While highly suitable for certain goals, the dataset is much too detailed for general use and widespread publication. Statistics Iceland follows strict policies aimed at guarding personal information and does not publish any data wherein individuals might easily be identified. A dataset that literally draws a line between a person’s home and workplace clearly falls beyond those guidelines.
In order to overcome this issue, Statistics Iceland employed small statistical units to aggregate individuals to ensure the masking of personal information. This abstracted network should be thought of as a companion to the more precise routing network, providing summary statistics and showing commuting rhythms at a broader scale in a format that can be published without restrictions.
This presentation outlines the technical aspects of generating the commuter network as well as comparing the two outcomes. It ends by considering how this type of NSO product might help inform policy and understanding of society.