In “Policy by Simulation” TNO and IAS join forces to unravel complex challenges; trying to understand what kind of interventions will produce what type of outcomes. Multiscale modelling of complex systems forms the core of the collaboration. The one year celebration included a workshop on the strategic development of the initiative, and presentations of student projects in various domains.

Presented student projects:

How do different policies influence the demand for car sharing?

Student: Bram van Duijvenboden
Supervisors: Lera Krzhizhanovskaya (UvA), Quinten Meertens (UvA/CBS), Ton Bastein (TNO)

Roads in and around cities are jammed, the air quality in some streets is sickening, and more and more public space is consumed by unused cars. Car sharing is promoted as one of the solutions for these urban problems, but the dynamics that determine demand and its implications for these problems are hardly known. This thesis applies system dynamics modelling to assess the demand for car sharing. Five policies that aim to promote car sharing are simulated. Four of these policies are: decreasing the cost of car sharing, increasing the costs of a private car, creating parking restrictions, and increasing the number of car sharing vehicles. As a fifth policy, the three most successful policies are combined. Simulation results show that: The number of car sharing vehicles heavily restrains the increase of car sharing. Without a proper public transport system, the car sharing demand comes mostly from public transport, and not from private cars. Most importantly, combining different policies is necessary to decrease the number of private cars.

The impact of social media on (sustainable) consumption and policy of Dutch supermarkets

Student: Joeri Athmer
Supervisors: Cees Diks (UvA), Quinten Meertens (UvA/CBS), Ton Bastein (TNO)

Social pressure and the prominent role of social media nowadays motivate the research question of investigating whether social media has an impact on (sustainable) consumption. To examine this potential influence, we linked four time series derived from the considered social media data with the weekly fraction of the turnover of sustainable meat and fish products over the total turnover of meat and fish products of two supermarket chains by the estimation of bivariate models. The model selection procedure, based on several statistical tests, resulted in solely VAR models. For these estimated models we examined the potential presence and direction of Granger causality.

We found that for certain variable combinations there is statistical evidence for Granger causality. Hence, on the basis of some of the considered variables it appears that social media influences sustainable consumption. We, however, stress that we have to be prudent with drawing conclusions based on this investigation. This is, in addition to various segments that have to be further investigated, caused by the complexity of the many competing modern life distractions for consumers in a supermarket and consumer environment. However, we believe that this thesis serves well as starting point on a rather untouched topic.

Opponent organisation modelling

PhD Candidate: Koen van der Zwet
Supervisors: Peter Sloot (UvA), Tom van Engers, (UvA/TNO), Ana Isabel Barros (TNO)

Terrorists, insurgents and criminal organisations yield a threat to the stability of societies and endanger democracy and peace. In order to develop efficient law enforcement strategies, we need to deepen our insight in the behaviour of such opponent organisations. This is rather challenging, since the covertness of their illicit operations hinders the possibility of extensive and detailed empirical research in actual networks. To overcome this difficulty we aim to acquire insight in the complexity of the emergent behaviour of opponent organisations by conducting multidisciplinary analysis of the behaviour of individuals. We have created an agent-based model that allows us to experiment and analyse the effects of various law enforcement policies at the opponent organisation level.