As a postdoctoral researcher, you will take primary responsibility for the methodology and software development at the heart of COMTIP. You will develop and refine agent-based, network, and dynamical-systems models – capturing tipping points, hysteresis, cascades, and reversible dynamics – and pair them with information-theoretic and causal tools to locate driver nodes and leverage points: the places where small interventions tip a system, and where they can backfire. You will turn this into well-documented, reusable open-source software and reproducible pipelines, and help shape a visual science–policy interface that communicates model outcomes to policymakers. The energy transition is the first application testbed – how incentives and policy can stimulate sustainable behaviour, starting with technology adoption – but the methodological core is domain-general, with scope for a second use case in year two. You will work closely with a PhD researcher (employed at RIVM) on healthy food choices and lifestyle-related diseases, integrating qualitative inputs (group model building, causal loop diagrams) with quantitative data, and you will help bridge complexity-science knowledge from the UvA into RIVM. You are expected to take an active role in team activities, from seminars and stakeholder workshops to collaborative papers and grants.

Tasks and responsibilities:

The list of tasks is not exhaustive and covers aspects of the project that may or may not be taken on by the Postdoc. If there are skills you already have or would like to develop, or a topic you are particularly interested in, please mention these in your motivation.

• Develop and refine agent-based, mean-field, network, and dynamical-systems models of collective behaviour, capturing tipping points, hysteresis, cascades, heterogeneous preferences, and reversible dynamics
• Apply information-theoretic, network, and causal tools to detect tipping points and identify causal driver nodes and leverage points — where, when, and how an intervention shifts (or backfires on) collective behaviour
• Formulate and validate computational methods to detect, analyse, and test social tipping points, and define when a model is “fit for purpose” for diffusion and intervention analysis
• Design and maintain well-documented, reusable, preferably open-source scientific software and high-performance simulation tools (e.g., parallelised agent-based models) with reproducible pipelines
• Help shape a visual science–policy / simulation interface that communicates model outcomes to policymakers and stakeholders
• Integrate diverse data sources and local contextual factors (economic, social, urban) with psychological factors, combining quantitative data with qualitative insights from group model building and causal loop diagrams (with room to develop these skills if they are new to you)
• Contribute to the design and analysis of discrete-choice (conjoint) experiments and survey measures that parameterise and validate the models
• Demonstrate the framework on the energy transition testbed (incentives and policy for sustainable technology adoption), with the option of a second use case after year 1
• Collaborate with the RIVM PhD researcher and facilitate reciprocal knowledge exchange between POLDER/IAS and RIVM, including a secondment at RIVM of at least one day per week
• Publish in peer-reviewed journals, present at international conferences, and engage in science communication and public outreach
• Contribute to team activities at CSL and POLDER/IAS (workshops, seminars) and collaborative papers and grants
• Stimulate the collaboration with RIVM
• Assist in teaching and supervision (e.g., tutorials, MSc/BSc projects) in consultation with the group, within the MSc Computational Science and MSc Complex Systems and Policy programmes

• A completed PhD in computational science, (computational) social science, complexity or network science, computer science, physics, applied mathematics, or a closely related field
• You have experience translating behavioural responses into computational models
• You can translate messy real-world questions into clear model assumptions and testable mechanisms, with a focus on explaining why a pattern arises, not only predicting it
• Demonstrable experience with quantitative analysis and scientific programming (e.g., Python and/or R; compiled or lower-level languages such as C++, Cython, or Nim are a plus)
• Demonstrable open-source software experience, or evidence of maintaining large software repositories, is a strong plus.
• You communicate clearly in spoken and written English and can write publishable academic texts
• Understanding of Dutch is a plus.
• You collaborate well in an interdisciplinary team and can work effectively with policy stakeholders
• Domain experience in the energy transition is welcome but not required: the role is methodology-first, and strong tipping-point and complex-systems expertise from any application area (e.g., health, ecology, governance, social or biological systems) transfers directly
• It is a plus if you additionally have experience with agent-based / diffusion modelling, information theory, network science, dynamical systems, complex contagion / threshold models and hysteresis, causal discovery, simulation-based inference, computational economics, discrete-choice modelling, group model building / causal loop diagrams, and/or data visualisation and science-policy interfaces

We offer a temporary employment contract for 38 hours per week for a period of a year with a probationary period of two months. The preferred starting date is November 1st, 2026, but it can be discussed.

The gross monthly salary, based on 38 hours per week and dependent on relevant experience, ranges between € 3,546 to € 5,538 (scale 10) This does not include 8% holiday allowance and 8,3% year-end allowance. The UFO profile Onderzoeker 4 is applicable. A favourable tax agreement, the ‘30% ruling’, may apply to non-Dutch applicants. The Collective Labour Agreement of Universities of the Netherlands is applicable.

Curious about our extensive secondary benefits package? You can read more about it here.