Network analysis on Persons for Official
Statistics
The census, which is as old as
civilization, is the origin of Official Statistics. It measures various sociodemographics
properties of the population which were, are and will be of importance for social scientists,
historians, policy makers and government.
A census is a very valuable source of information but its description of society formed by the
connections between people is very limited. It records people living on the same address,
but it fails to
capture the broader network of relationships: family, friends,
neighbors, co-workers and acquaintances. Most demographic statistics describe
(aggregates of) properties of inhabitants. If Official Statistics strives to measure society,
describing the network of relations between people that form
the fabric of society can be a source of interesting demographic statistics.
Is the strength of family ties regionally correlated? How diverse are personal
networks? Given the current demographic trend in most countries that the
average age is increasing, do parents live close to their children or is this
distance increasing?
At Statistics Netherlands a research program was formed that tries to use complexity science
and network analysis to derive new and additional official statistics. Several projects are
started including deriving a enterprise to enterprise network for describing economic networks,
but also a project to derive a social network of the Netherlands.
This abstract describes the derivation of a directed network
with family, dwelling, neighbor, school going children
and coworkers relationships for the 17 million inhabitants of the Netherlands and some of its
potential use for producing official statistics from it.
Reference:
IPS03-001
Session:
Data Exploration and Visualisation
Presenter/s:
Edwin de Jonge
Presentation type:
Oral presentation
Room:
GASP
Chair:
Giuditta DE PRATO, European Commission - Institute for Prospective Technological Studies, (Email)
Date:
Tuesday, 12 March
Time:
17:15 - 18:15
Session times:
17:15 - 18:15