Complex problems call for a complex approach, new UvA research initiative says
Developing computer models that allow policymakers to devise appropriate solutions to complex issues such as climate change. That’s the goal of Polder, the UvA research center that officially opened this week.
“Complexity and interdisciplinarity don’t fit well with the practice of policymakers,” Albert Faber of the Scientific Climate Council revealed halfway through Polder’s opening meeting in the presence of around 40 attendees, including some policymakers but mostly scientists, at the Science Park. Polder is the research initiative of the Institute for Advanced Study (IAS).
Yet that is exactly what Polder wants to achieve. The center wants to provide policymakers with tools from complexity theory, a relatively young branch of science that describes systems that are difficult to predict because of all sorts of relationships and dependencies to find solutions to complex problems.
Nitrogen, climate change, and the housing crisis, for example, are now too often treated as separate problems when in fact there is much overlap. Polder wants to change that by collaborating with as many disciplines, companies, and organizations as possible and by taking human behavior into account. So in the case of nitrogen, the computational model of complexity theory would cover not only nitrogen but also farmer behavior and groundwater levels.
What-if scenarios
Polder works according to a clear roadmap. In group model-building sessions, stakeholders sit down together and sketch with pen and paper how the various contributing factors to a problem are related. After several rounds, the process yields a qualitative model that can sometimes already provide solutions.
For a more thorough approach, that model is translated into a computer model and additional research is conducted. Based on that model, policymakers can sketch "what-if" scenarios in a virtual world to arrive at the right interventions. Polder focuses primarily on health and sustainability and has already collaborated with the University Medical Center, the national police, several ministries, and other organizations.
Polder is also working on a new two-year master’s program, “Complex Systems and Policy,” to be launched in September 2025. This interdisciplinary master’s will combine data science and modeling techniques with skills from behavioral science, governance, and policy.
Post-its
During the opening at Science Park, attendees get a taste of Polder in a mini-workshop on modeling. The case study: the energy transition to electric cars. Participants affix post-its indicating factors influencing that process such as costs, subsidies, and the availability of charging stations on a large sheet of paper. The participants then indicate the connections between the post-its with plus and minus arrows.
It soon becomes clear: This is difficult. All sorts of discussions ensue. “Do lower electric car costs always lead to more users? Might the car lose its function as a status symbol?” and “If the municipality tightens the regulations for charging stations, will people then get rid of their electric cars?”
It also quickly becomes clear that the choice of people at the table rather determines how the problem is outlined. After just under an hour and six post-its later, the group sits there somewhat defeated. “Give us another 100 hours and we might come close,” sums up one of the participants.
“Building models is indeed a time-consuming process,” program manager Ted Jan Post says afterwards. “More time is also normally taken for that in several rounds. After the pen-and-paper phase, the model must also be quantified to become a computer model. That can sometimes take years. But it is not always necessary to go into so much detail. We therefore deliberately work in a result-oriented way because urgent challenges such as health care and climate cannot wait that long.”
