Dutch Computational Science Day (DUCOMS) 2025

Computational Social Science at Scale

The social sciences may, traditionally, not be associated with large-scale computing, in the way that for instance astrophysics and particle physics are. Perhaps the implicit assumption is that the only data available to social scientists concern small-ish groups of people (fewer than a few thousand individuals) having been surveyed, and the only computing being applied is to do some statistical regressions or association tests.

In fact, in the Netherlands large datasets with a very broad range of variables covering many aspects of the entire population (by now 18+ Million people) are available for research, already for several decades (albeit under certain strict conditions due to privacy regulation). The most recent 5 years, and in particular the covid-19 pandemic, also highlighted the need for fine-grained behavioural analyses: to model and measure how people move around, and who they meet, and how this affects contagion risk. For rather longer, although not as much present in public perception, fine-grained modelling has been done to assess the economic effects on household budgets of government policies, usually just before every year in September the national budget is presented in parliament.

In my talk I will present a brief overview of recent work in social network science. I will attempt to cover the mapping of the several billion social ties between all residents in the Netherlands, and the use in social studies on eg. segregation. I will also pay some attention to agent-based modelling, the social sciences equivalent of particle dynamics, to model the interactions between the roughly 18 million residents of the Netherlands. Often, such simulations are calibrated using the large databases of Statistics Netherlands, to study social and economic behaviour.

Frank Pijpers
CBS & University of Amsterdam

About Frank Pijpers

Frank Pijpers has worked at the CBS since 2010, as a senior methodologist since 2014, and, since 2015, as a programme manager for methodological research. Previously, he was a government operational researcher (senior) in the Department for Children, Schools and Families of the UK government, a lecturer at Imperial College London (United Kingdom) and associate professor at Aarhus University (Denmark). Pijpers studied Physics and Astronomy at Leiden University and obtained his PhD at the same university for his thesis entitled The Dynamics of the Winds of Cool Giants and Supergiants. Since 2021, Frank Pijpers is professor by special appointment of Complexity for Official Statistics at the Faculty of Science at the University of Amsterdam (UvA). The chair was established on behalf of Statistics Netherlands (CBS). Pijpers combines his professorship with his functions as programme manager for methodological research and senior methodologist at the CBS.

Would you like to present your work at DUCOMS?