A murmuration of birds
© Hollandse Hoogte

A digital twin of society illuminates its complexity

The statistical studies by Statistics Netherlands (CBS) mirror Dutch society. But what causes the emerging trends identified in these studies such as the patterns of the economic cycle or social mobility? To help answer this question, Prof. Frank P. Pijpers of Statistics Netherlands will simulate a 'digital twin' for every resident of the Netherlands. This computer simulation with 17.5 million people, is to mimic the complex system of Dutch society and help to understand it better.

A collaboration agreement signed between Statistics Netherlands and CERN – the largest particle physics laboratory in the world – foresees the usage of BioDynaMo technology codeveloped at CERN, to construct a ‘digital twin’ of the Netherlands. BioDynaMo is a software platform that can be used to virtually simulate the behaviour of different ‘particles’ in a ‘system’ where they all influence each other.
To understand the concept of the ‘digital twin’ better, let us take the example of a flock of starlings, known for forming impressive flight patterns against the evening sky. These patterns arise while an individual starling in that swarm knows nothing about it. It makes its own choice and only pays attention to neighbouring peers.

People also make their own choices. They take more things into account than starlings do, yet the social trends that emerge as a consequence, are not self-evident. For example, different groups of people may live side by side in the same cities, but rarely or never meet or speak to each other, without making that conscious choice. As a result, their opinions, and habits may also diverge further and further from each other.

By applying modern simulation techniques such as agent-based modelling to the data from their System of Social Statistical Files (SSB), Statistics Netherlands can visualise the collective behaviour of groups in Dutch society. One example would be the segregation of the population into groups with diverging lifestyles, based on the behaviour of the approximately 17.5 million individual inhabitants. The aim is to investigate, among other things, which mechanisms are responsible for maintaining or increasing inequality between population groups and across generations in areas such as education, economic and social opportunities.

The insights provided by the computer simulations of the ‘digital twin’ of the Netherlands are important for planning bureaus and ministries. Based on these insights, they can, for example, develop more effective policies to combat inequality in line with the United Nations' sustainability goals. For Statistics Netherlands it is invaluable to offer a good model that has been tested for validity and applicability. With its help, planning agencies can better assess how policy choices affect the population and how people adjust their behaviour based on the policy. This also makes unintended consequences of policy more visible.

Prof. Frank P. Pijpers is a researcher at Statistics Netherlands and professor by special appointment of complexity for official statistics at the University of Amsterdam.

Read the story as covered by CERN’s Knowledge Transfer here.