CBS and UvA in collaboration to solve social problems

/ Author: Masja de Ree
Formal signature of the collaboration between CBS and UvA/IAS
© Sjoerd van der Hucht Fotografie / My Eyes4u productions
It sounds like science fiction: a ‘digital twin’ of society that policy-makers can use to resolve complex social problems. But that’s exactly what researchers at Statistics Netherlands (CBS) and the University of Amsterdam’s (UvA) Institute for Advanced Study (IAS) are working on right now. This collaboration between CBS and UvA has already produced positive insights, and on 31 May it was extended for a further five years.
We live in a world of complex systems: everything is interlinked, from the economy and the environment to social security and the way people engage with their network of friends, colleagues and acquaintances. A whole range of factors determine how these systems develop, and of course the systems also influence each other. Since 2019, researchers from CBS and UvA/IAS have been studying that complexity and trying to get to grips with complex problems such as climate change and inequality of opportunity.

Getting to grips with complexity

The collaboration has two objectives. ‘In the first place, we want our study to contribute to solving complex social problems such as the poverty trap and the energy transition,’ explains IAS Director Prof. Huub Dijstelbloem. ‘Secondly, the collaboration between IAS and CBS is helping us gain a deeper understanding of complexity as a method. We are especially interested in how different complex systems are related, for example the relationship between the environment, poverty and inequality. Our goal is to develop a better method for researching systems, as well as to better understand those systems so policy-makers can make targeted improvements.’

Digital twins

A lot of research has been done in recent years into what is known as ‘digital twins’. As CBS researcher Gert Buiten explains, ‘You can research complex systems by creating a digital copy of reality. In effect, you build a copy of Dutch society and insert “agents” to represent the people who live in the Netherlands. Then you can model the behaviour of those “agents”, and with the push of a button you can initiate different scenarios and watch how they unfold.’ Dijstelbloem explains: ‘Policy-makers can use those “digital twins” of a society to help them plan targeted improvements to that society. For example, a simulation in a digital version of society can flag up early warning signals, and you can test options for new policies.’ This research method raises many questions about ethics and privacy, as CBS researcher Edwin de Jonge is well aware. ‘Those aspects are always on our mind. We ensure privacy by design: privacy safeguards are baked into the design of our models.’

Researchers of CBS and UvA/IAS talk to each other to share new ideas and plans

Studying networks

One next step is for CBS and the IAS to look into networks of relationships between people. CBS has used information contained in government registers to monitor relationships among the Dutch population, whilst maintaining total anonymity. ‘We’re talking about family relationships,’ says De Jonge, ‘who is living with whom, which people know each other through work, are friends or are at the same school?’ This database has already been used to investigate the extent to which highly educated Dutch people come into contact with their less well-educated compatriots, with citizens’ privacy guaranteed. The network has also been used to calibrate scientific models for the spread of coronavirus, and to make them as realistic as possible. ‘Researchers call this a unique database,’ Buiten contributes. ‘Together with the IAS, we will devise even more applications for this resource in the coming years. We also work with Leiden University, in a consortium with the IAS, to conduct Population Scale Network Analysis (POPNET) to create a secure user interface for the database. Under strict conditions, this enables researchers from ministries and public policy and research institutes to also draw on this resource.’

Persuading people

Dijstelbloem also stresses a desire to contribute to ‘the narrative’. ‘With really complex problems, such as climate change or the coronavirus pandemic, you often see that developments don’t come in a logical order. Instead of gradually worsening, the situation can suddenly escalate or reach a point of no return. That means you have to act before it’s too late. A government may need to take measures that go against people’s intuition, such as imposing the coronavirus restrictions. Then you have to persuade people and show them that they’re acting for the greater good, for society as a whole. At the IAS we want that to be done the right way, so we work with politicians and policy-makers to help find ways to explain things clearly.’

New ideas and plans

To coincide with the extension to the collaboration, a workshop was organised to stimulate researchers to share new ideas and plans with each other. As Buiten describes it, ‘There’s a lot of enthusiasm, it’s buzzing.’ There is clearly a great deal still to come. Dijstelbloem shares an example: ‘We’ll be focusing on the issue of polarisation in the near future. We aim to use resources such as CBS’ demographic and economic data to gain a better understanding in that area. Where do people with more extreme opinions live, for example, and how are they distributed across the labour market? We already do that through surveys, for example in schools, but we want to do more. So we’re investigating ways to better understand this complex social phenomenon.’

Combining theory and practice

‘At the same time,’ De Jonge adds, ‘we’re continuing to work on our “digital twins” of Dutch society that combine agent-based models with our network database.’ Such a comprehensive system demands a lot of calculating power, so to secure that power CBS is working with the CERN research institute in Switzerland. ‘We combine their calculating power with our real-life data and the IAS’ mathematical models,’ Buiten explains. ‘It’s opened up amazing opportunities. It’s a bit like the James Webb telescope: our models give you a new perspective on the world that just wasn’t possible before. That generates new theories and new scientific insights, and hopefully also solutions to our biggest social problems.’