The role of causal modelling in official statistics
Causal modelling is of interest to augment the interpretation of data published by national statistical institutes. An exploration of the relevant methods as well as a case study in the domain of education are the main focus of this paper.
This paper is intended to explore to what extent causality research has a role in innovative output of Statistics Netherlands. In the more appropriate, cautious, approach, the potential of using data to exclude putative causal mechanisms appears to be the most promising, and is an active line of research. A limited proof-of-concept study on secondary school selection procedures is explored as an example of the technique but should not be interpreted as exhaustive research on counterfactual causality or counterfactual fairness within school level selection in the Netherlands.odel has been extended with effects of the Covid-19 pandemic on mobility in 2022, besides those in 2020 and 2021.
